Specificity of the Diagnostic Materials for

Laryngopharyngeal Reflux

In fulfilment of the requirements for the degree

MA Speech Pathology by research

Jeanneane Logan

0112716W

February, 2011

Declaration

I hereby declare that this dissertation is my own unaided independent work. It has not been submitted before for any degree or examination at this or any other academic institution, nor has it been published in any form. February, 2011 Jeanneane Logan 0112716W

i

Acknowledgements

The researcher would like to extend her thanks to the individuals and agencies that

were instrumental in making this research possible.

Firstly, I would like to extend my sincere gratitude to Mrs. Heila Jordaan, my research

supervisor, for her help, insight and dedication. Her enthusiasm and support during

my masters has encouraged me to extend myself and fuelled my interest in the area of

voice. Thank you for your long hours and patience during this process and ultimately

thank you for

To Dr. Lance Maron, thank you for allowing me unlimited access to your patient files

and clinic. Thank you for giving me permission to conduct my research project in the

Donald Gordon Voice and Swallowing clinic and for accommodating me on many

days and afternoons in your clinic. Without your and your staffs input and cooperation

this research would not have been feasible.

Thank you to my family for their support and endless cups of tea during the write up

of my data and results. To my father who has assisted me with my research by proof

reading and discussing ideas with me. He has in the process learned more about

Laryngopharyngeal reflux than he ever imagined he would.

Thank you to my husband, Ryan, for his love and support. Thank you for

understanding many late nights, weekends, early mornings and holidays for

supporting me and providing all the assistance you could where possible.

iii

Table of Contents

Acknowledgements i Abstract ii Introduction and Rationale pg. 1-4 Literature Review pg. 5-24 Method

Aims pg. 25 Research Design pg. 25-26 Participants pg. 26-34 Ethical Clearance pg. 34 Measures pg. 35-37 Procedure pg. 37-38 Data Analysis pg. 39 Reliability and Validity pg. 40

Results

Description of Results on the RFS and the RSI pg. 41-45 Trends on RFS and RSI rating pg. 45-57 Correlation of the total RFS and RSI ratings, pg. 57-59

and inter-item and intra-scale correlations RSI intra-item correlation pg. 59-62 RFS intra-item correlation pg. 62-64 RFS and RSI inter-item correlation pg. 64-66 Effects of gender, age, smoking and professional pg. 67-69

voice use on the RFS and RSI totals Trends related to smoking, gender, age pg. 69-75

and professional voice use Discussion and Conclusion

Interpretation of Results Previous Studies pg. 76-77 Current study pg. 78-86

Trends pg. 86-92 Conclusion pg. 93-96 Limitations of the study pg. 97-98 Implication of the research pg. 98-100 Direction for future research pg. 100-102

Reference List pg. 103-113 Appendices

Appendix A-Patient consent form pg. 114-115 Appendix B-Reflux Severity Index (RSI) sheet pg. 116

iv

Appendix C-Reflux finding score (RFS) sheet pg. 117 Appendix D-Letter of permission from research site pg. 118 Appendix E-Ethical Clearance Certificate pg. 119

Tables

3.1 Description of patients pg. 31 3.2 Measures of Central Tendency of patient’s age pg. 32 4.1 Summary statistics for the RFS and the RSI pg. 41 4.2 Exact rating on the RFS and number pg. 43

of participants to whom each rating applied 4.3 Number of participants to whom each rating applied pg. 45 4.4 Total for each sign of the RFS and percentage for each total pg. 46 4.5 Total for each sub-rating of the RSI pg. 48

and percentage for each rating 4.6 Correlation matrix for the RFS and the RSI pg. 58 4.7 Intra-item correlations on the RSI between pg .59

hoarseness and other items 4.8 Intra-item correlations on the RSI between pg. 60

throat clearing and other items 4.9 Intra-item associations on the RFS pg. 63 4.10 Inter-item association between the RFS and the RSI pg. 65 4.11 Summary statistics for the RFS and the RSI by gender, pg. 67

smoking status and professional voice use 4.12 Mean ratings for each age group on the RFS and the RSI pg. 68

Diagrams

3.1 Flow diagram representing the procedure for data collection pg. 30 3.2 Pie chart illustrating the relationship between smoking pg. 33

status and professional voice status 3.3 An example of the coding system utilised in pg. 38

the excel spreadsheet 4.1 Number of participants who attained various RFS scores pg. 42 4.2 Number of participants who rated the total RSI as pg. 44

various severity scores 4.3 Number of participants for each rating on hoarseness pg. 49 4.4 Number of participants for each rating of Mucus pg. 50 4.5 Number of participants for each rating of Throat Clearing pg. 51 4.6 Number of participants for each pg. 52

rating of Lump in the throat 4.7 Number of participants for each rating of Heartburn pg. 53 4.8 Number of participants for each rating of Annoying Cough pg. 54 4.9 Number of participants for each rating of pg. 55

Coughing while Lying Down 4.10 Number of participants for each rating of pg. 56

Swallowing difficulties 4.11 Number of participants for each rating of pg.57

Breathing difficulties

Introduction and Rationale

Laryngopharyngeal reflux is defined by Sataloff, Hawkshaw and Gupta, (2010), as “an

extraoesophageal variant of gastroesophageal reflux disease, (GORD), that affects the

larynx and pharynx” (p 124). Hoarseness, chronic unexplained cough, frequent throat

clearing, and a feeling of a lump in the throat are all common symptoms of

laryngopharyngeal or “silent” reflux (Fouad & Rifaat, 2010). The general nature of these

symptoms translates into difficulties with differential diagnosis and at present, the

diagnosis and treatment of extraoesophageal reflux disease is still controversial (Lee,

Kim, Ryu, Kim, Cheong, Lee & Song, 2010).

“There is no consensus on the diagnosis and treatment of laryngopharyngeal reflux and

the majority of clinicians depend mainly on clinical findings and empirical therapeutic

tests rather than more specific investigations.” (Ali, 2008, p. 28).

In fact the mere existence of ‘silent reflux’ is currently controversial (Barry & Vaezi,

2010). Due to such controversy and lack of definitive diagnostic criteria

laryngopharyngeal reflux remains challenging to diagnose and manage effectively.

The aim of this study is therefore to examine the specificity of, and relationship between

diagnostic methods used in laryngopharyngeal reflux, (LPR), disease, specifically the

Reflux Symptom Index, (RSI), and the Reflux Finding Score, (RFS). Through a

retrospective chart review of 105 patients this research attempted to provide a degree of

clarity regarding the controversial nature of LPR.

As mentioned, one of the main symptoms of LPR is a hoarse voice. A complex interplay

of muscle contraction, inhalation, relaxation and exhalation in a sequenced and timed

manner allows a vibratory movement that is audible as voice (Tanner, 2007). This

vibration is unique to each person. Voice is essentially part of an individual’s personality.

For many it defines who they are. Consider for example how we recognise people on the

phone, at the door or on the radio by their voice. Voice is used as a means for nonverbal

communication (Tanner, 2007). We make judgments about people based on their voice,

such as their age, gender, educational level and even in some instances socioeconomic

2

background. Voice is strongly linked to emotion and is used to convey primary emotions

such as: anger, fear, joy, disgust and surprise (Titze, 1994). Personality traits of

confidence, kindness, honesty and cheerfulness are associated with vocal characteristics.

If an individual’s voice is altered it may therefore affect emotion and personality.

For many, voice comprises a crucial and irrevocable part of their personal and

professional lives. Modern technology has also come to rely on voice more and more

with vocal recognition dialling, security systems and voice-to-text typing (Tanner, 2007).

Consequently, fluctuations in voice quality may be insignificant or may be catastrophic

and cause a large amount of suffering in individuals. For some it may lead to changes in

vocation, lifestyle, hobbies, social patterns and ultimately quality of life (Tanner, 2007).

Professionals who use their voice in order to work e.g. lawyers, ministers, teachers,

singers, sales people and speech therapists - may be unable to work if their voice is

affected even mildly (Titze, 1994).

Voice disorders may reduce job effectiveness and productivity and may even lead to job

loss in many professions. Vocal disorders are evaluated and may be classified as

disabilities which can be a huge cost to company. People who suffer from voice disorders

may be candidates for vocational rehabilitation and they may require compensation for

lost earnings and retraining for a new job (Tanner, 2007). In many instances people may

need to be put on extended sick leave and often are unable to return to their previous

profession due to their poor vocal quality.

A study by Roy, Merrill, Grey and Smith, (2005), of 1326 people randomly selected from

the general population in the United States revealed that 29,9% of the sample had

suffered from a vocal pathology during their lifetime. Of the sample, 4,3 % indicated that

their job performance had been negatively impacted by their voice disorders and had

hindered and obstructed their ability to perform certain job tasks. Attendance at work was

also cited as being impeded by voice problems. Further 7,2% of participants reported they

were absent from work for 1 or more days in the previous year and 2% reported more

than 4 days of voice-related absence. (Roy et al., 2005).

Another study found similar results with teachers who have voice disorders. Results

indicate that teachers with vocal disorders go into retirement earlier than those without

3

vocal disorders and are more likely to have to change professions compared to their non

voice disordered counterparts (Chen, Chiang, Chung, Hsiao & Hsiao, 2010). Voice

difficulties therefore have an impact on attendance at work and ultimately on

productivity.

Vocal disorders are a disability and create a feeling of helplessness, grief, anguish and

loss. Therefore disorders such as laryngopharyngeal reflux that create vocal symptoms

needs to be diagnosed and treated effectively in order to reduce these and other effects on

people.

Voice disorders are generally treated by a team of specialists, with the speech therapist

making a significant contribution. The speech therapist’s role is to provide behavioural

treatment for the vocal pathology and aid the patient in managing the vocal symptoms

with which they present, through a number of procedures and techniques (Colton &

Casper, 1996). The speech therapist also aims to educate the patient regarding: the vocal

mechanism, identification and reduction of phono-trauma and high-risk vocal situations,

conservation of voice or vocal rest, controlling the amount of talking, monitoring vocal

pitch and intensity, encouraging local lubrication and systemic hydration, and promoting

optimal dietary considerations (Behlau & Oliveira, 2009). Many of these are intended to

aid in reducing further damage to the vocal folds by modifying certain behaviours e.g.

vigorous throat clearing.

LPR is found in a high percentage of patients who have laryngeal carcinoma (Galli,

Cammarota, Volante, De Corso, Almadori, & Paludetti, 2006). This is thought to be due

to inflammation and cellular mutation from direct chemical assault in the form of reflux.

LPR therefore represents a potentially life threatening disorder and it is crucial to have

precise and early diagnosis.

The findings outlined here indicate that the need for clarity in diagnosis and effective

management of LPR has become a priority. “Evidence confirming the diagnostic

significance of various complaints and findings is scarce and contradictory” (Sataloff,

Castell, Katz & Sataloff, 2006 p. 58).

4

Sataloff, Hawkshaw and Gupta (2010) outline the need for additional multidisciplinary

research into the diagnosis and treatment of LPR. If a more reliable method of diagnosis

could be established then LPR would not be such an enigma and appropriate treatment

could commence. Interdisciplinary studies should focus on trying to answer crucial

questions regarding sensitivity and specificity of the numerous findings associated with

LPR (Sataloff et al., 2006).

The Reflux Severity Index (RSI) (Belafsky, Postma & Koufman, 2002), is a useful tool in

drawing attention to the symptoms associated with reflux. Additionally, the relationship

between symptoms, signs and reflux severity needs to be characterised (Noordzij, Khidr,

Desper, Meek, Reibel & Levine, 2002). The Reflux Finding Score (RFS) (Belafsky,

Postma & Koufman, 2001b), and RSI scores are designed to grade severity and clinical

findings in LPR. Therefore, research into the relationship between the RSI and the RFS is

warranted and findings may have significant clinical relevance and impact.

5

Literature Review

The purpose of this literature review is to define core concepts and theories related to

laryngopharyngeal reflux with particular emphasis on appropriate diagnosis. The

rationale for the proposed research is also further developed.

Voice disorders account for a large proportion of the problems encountered by patients

seen at otolaryngology practices. A team approach is used to treat voice disorders as they

often require lifestyle, behavioural and medical intervention (Sataloff, 2005). The team

members involved with voice disorders would typically include a speech therapist, an

otolaryngologist and in cases of professional voice users a voice and/or singing coach

(Tanner, 2007). Each member of the team brings a particular set of skills and knowledge

to the diagnostic process. The otolaryngologist is highly skilled in the assessment of

health and disease in the larynx, whereas the singing coach recognises problems in vocal

technique that are specific to the singing voice (Colton & Casper, 1996). The speech

therapist’s role in vocal pathology is to diagnose the type of vocal disorder the patient

presents with. This is done through assessment of perceptual characteristics, case history

and patient interview where aspects such as respiration, prosody and tone are considered.

Patients are diagnosed and provided with an in depth individualised treatment plan to aid

in reducing vocal disorders and change damaging and pathological vocal behaviours. In

general the goal of voice therapy is to restore the best voice possible to aid in functional

communication and employment (Colton & Casper, 1996). The speech therapist is able to

recognise the potential for serious disease and the need for medical intervention and refer

appropriately. The speech therapist specialises in the phonatory function of the speech

mechanism and the manner in which various conditions disturb this (Colton & Casper,

1996).

Voice disorders have an impact on the quality of life and often negatively influence an

individual’s functioning. Both leisure and professional aspects are affected. The Voice

Handicap index (VHI), (Jacobson, Johnson, Grywalski, Silbergleit, Jacobson, Benninger

& Newman, 1997), quantifies this impact. The voice handicap index examines three

6

areas: physical (perceived vocal functioning and laryngeal discomfort), emotional

(emotional response to vocal problems) and functional (the impact of voice difficulties on

daily activities). Each aspect is assessed by ten questions rated by the patient on an

ordinal scale from 0 to 5 (Welham, Dailey, Ford & Bless, 2007).

A number of studies have examined patients’ subjective experiences of voice disorders

using the VHI (Maertens & de Jong, 2007; Rosen, Lee, Osborne, Zullo & Murry, 2004;

Orlikoff & Baken, 1993). A study by Chen, Chiang, Chung, Hsiao & Hsiao (2010), on

teachers with voice disorders indicated they had increased absenteeism, changes in

teaching style, changes in job performance and pressure in the functional sphere of their

lives. Further, they indicated a reduction of social activities overall and an avoidance of

social situations. They stated that the voice problem brought on feelings of

embarrassment, being upset and a negative effect on overall emotional state, with an

adverse impact on their self image. Patients have reported an increase in anxiety,

psychological distress and 77% of patients also reported reduced social activities

compared to controls without a history of voice difficulties (Siupsinskiene, Adamonis &

Toohill, 2007). The study further indicated that patients’ quality of life is more impacted

by symptoms of reflux than laryngoscopic findings would suggest (Siupsinkiene et al.,

2007).

The present study is an investigation of the symptoms and signs of a particularly common

condition, that both results in and compounds voice disorders namely laryngopharyngeal

reflux.

2. 1 Definition and incidence/ prevalence

Laryngopharyngeal reflux (LPR), is also known as pharyngeal, supra-oesophageal, silent

or extra-oesophageal reflux (EER), and refers to the backflow of stomach contents into

the laryngopharynx (Koufman, Aviv, Casiano & Shaw, 2002; Musser, Kelchner, Neils-

trunjas & Montrose, 2009; Belfasky, Rees, Rodriguez, Pryor & Katz, 2008). LPR is

7

closely associated with gastro-oesophageal reflux disease (GORD). The relationship

between LPR and GORD is further elaborated on in section 2. 2.

LPR is described as being extremely common and a potentially debilitating chronic

disease process (Aviv, Parides, Liu, Kaplan & Close, 2000). It has been reported that

more than 50% of patients complaining of hoarseness have LPR (Ford, 2005), and

further, that 4-10% of patients who, consult ENT’s do so for gastro-oesophageal reflux

disorder (GORD) complaints (Reimer & Bytzer, 2008). According to Ford (2005), 25%

of patients with LPR experience spontaneous resolution of symptoms while 50% have a

chronic course of intermittent remissions.

LPR can be found in all populations regardless of background. However, a trend has

emerged with the primary occupation of patients diagnosed with LPR being professional

voice users, especially singers (Sataloff, 2005a). This may assist the voice care team in

predicting which patients coming into voice clinics will present with LPR. Further, health

care workers may be able to treat such individuals prophylactically with lifestyle, dietary

and relaxation methods as well as with vocal hygiene measures.

Sataloff et al. (2006), postulate a number of reasons for this high prevalence of LPR in

professional voice users. Firstly, LPR may be due to the muscular mechanisms taking

place during singing where the stomach is compressed by forceful contraction of the

abdominal muscles. Secondly, singers often perform on an empty stomach and

consequently eat large meals before retiring for the evening. Thirdly, psychological stress

has been closely associated with gastroenterological conditions. Performance careers

have a high incidence of stress, which can increase production of stomach acid. Fourth,

many professional voice users consume a lot of caffeine and often make less healthy food

choices (Sataloff et al., 2006). The current study thus included a group of professional

voice users.

2. 2 LPR and GORD

LPR and GORD differ in symptomology, manifestations as well as in response to

treatment (Koufman et al., 2002). Patients diagnosed with GORD suffer primarily from

8

heartburn and oesophagitis, which is not the case in LPR (Reimer & Bytzer, 2008;

Schreiber, Garten & Sudhoff, 2009). GORD in general, tends to occur nocturnally in

supine position, whereas, LPR is an intermittent daytime occurrence, when the patient is

upright (Abaza, 2004). GORD is characterised by prolonged periods of oesophageal acid

exposure and dysmotility (Koufman et al., 2002). GORD is postulated to occur from

lower oesophageal sphincter dysfunction and dysmotility leading to prolonged periods of

oesophageal acid exposure (Koufman, Aviv, Casiano & Shaw, 2002). The

laryngopharynx is not exposed to acid for prolonged periods of time in LPR, (unlike the

oesophagus is in GORD), and the primary defect anatomically in LPR is upper

oesophageal sphincter dysfunction. (Koufman et al., 2002). It is likely that these

differences in anatomical dysfunction lead to the differing presentations of GORD and

LPR.

Another difference between LPR and GORD is that an increased body mass index (BMI)

or obesity is not positively linked to pharyngeal reflux in LPR patients whereas GORD is

strongly linked to increasing BMI and obesity (Halum, Postma, Johnston, Belafsky, &

Koufman, 2005). It is clear that LPR and GORD are not the same entity. Their aetiology

must be different considering the discrepancy between the symptoms and presentation of

the two disorders.

LPR and GORD are however, both due to mucosal injury from acid and pepsin exposure

(Remacle & Lawson, 2006). They differ in that the oesophagus has inherent structures

and processes that assist in preventing mucosal injury. These include bicarbonate

production, mucosal tissue resistance and oesophageal motor function with acid

clearance. The laryngopharynx by contrast does not have such protective mechanisms

(Remacle & Lawson, 2006). The laryngeal mucosa is fundamentally different to

oesophageal mucosa and is more sensitive to acid and pepsin (Hammer, 2009). The

larynx is more susceptible to reflux injury than the oesophagus, since it has fewer

extrinsic and intrinsic epithelial cells than the oesophagus (Koufman et al., 2002). Even

brief exposure to small amounts of reflux can cause damage to the laryngopharynx. Thus

the oesophagus is well equipped to deal with reflux whereas the larynx and

laryngopharynx is less so. The larynx is more delicate than the oesophagus with ciliated

respiratory epithelium. This epithelial structure enables mucus to be cleared from the

pharynx and trachea. Therefore when these structures are compromised through failure of

9

any of the protective mechanisms of the upper aerodigestive tract (upper and lower

oesophageal sphincters, oesophageal motor function and cellular construction of

oesophageal mucosal tissue), mucus stasis follows (Ford, 2005). The resultant

accumulation of mucus leads to a post nasal drip and throat clearing (Ford, 2005).

Choking and coughing can be caused by irritation of the airway due to a direct

mechanism of contact with the refluxate material and the reflux of gases into the

oesophagus (Hanson & Jiang, 2000; Mosca, Rossillo, & Leone, 2006). This combination

of factors can lead to vocal fold oedema, contact ulcers and granulomas that lead to

typical symptoms of hoarseness, globus pharyngeus (the sensation of a lump in the throat)

and sore throat reported by patients suffering from LPR (Ford, 2005). Globus sensation is

caused by cricopharyngeal muscle spasms and it is not necessary for the acidity to reach

the pharynx in order to bring on this symptom (Mosca et al., 2006). Inflammation and

irritation may in turn promote excess mucus production as a protective mechanism.

Four main mechanisms responsible for the presentation of LPR have been outlined by

Maron and Jordaan (2010). These are: chemical, mechanical, hypersensitivity and

immunologic.

The chemical mechanism is mainly due to the acid and non-acid components of the

refluxate: bile and pepsin which cause microscopic inflammation of the mucosal lining.

The refluxate dilates the intercellular spaces by breaking down cell membrane integrity

and constructing a more permeable environment which allows the refluxate to penetrate

the submucosa (Maron & Jordaan, 2010; Johnston, Knight, Dettman, Lively & Koufman,

2004).

Pain and heartburn are explained through the mechanical processes of oesophageal lumen

distention and contraction of the musculature. Some of the symptoms that present

clinically (coughing, lump in the throat etc.) are due to stimulation of the peripheral and

central vagal nerve fibres caused by hypersensitivity of the gastrointestinal tract (Maron

& Jordaan, 2010; Gupta & Sataloff, 2009). Bronchospasm, chest pain, cough and asthma

may be the result of acidification of the distal part of the oesophagus which stimulates

acid sensitive receptors (Maron & Jordaan, 2010). In addition, the refluxate may

indirectly stimulate the vagal nerve fibres in the oesophagus (Reimer & Bytzer, 2008),

10

leading to an ongoing and persistent cough which damages the vocal chords due to

continued contact.

The fourth mechanism is an immunologic phenomenon within the submucosa which leads

to inflammation. Papilla elongation, basal cell hyperplasia and dilated intercellular spaces

are all histological changes that take place in the presence of reflux (Maron & Jordaan,

2010).

LPR may also be associated with Paradoxical Vocal Fold Movement (PVFM), vocal

nodules, postnasal drip, allergies, sinus inflammation, and various pulmonary diseases

(such as asthma and chronic obstructive pulmonary disease) and polypoid degeneration

(the membranous portion of the vocal fold becomes filled with thick gelatinous fluid)

(Koufman et al., 2002; Barry & Vaezi, 2010). Lifestyle modifications such as cessation of

smoking, weight loss, and decreased activities that cause stress on the voice are advised

for these cases (Barry & Vaezi, 2010). According to Carrau, Khidr, Gold, Crawley,

Hillson, Koufman and Pashos (2005), a number of studies indicate that acid reflux is

present in 50-80% of patients with asthma, 10% to 20% of patients with chronic cough,

and 25% -50% of patients with globus sensation. This is in all likelihood due to the

inherent nature of LPR. Changes in the anatomy and physiology of the hypopharynx as a

result of LPR can also lead to dysphagia as well as a number of other symptoms (Aviv,

Liu, Parides, Kaplan & Close, 2000).

2. 3 Symptoms and complications of LPR

It is reported that the most common symptoms of LPR include throat clearing, persistent

cough, voice-quality change and globus pharyngeus (Ford, 2005; Remacle & Lawson,

2006; Sataloff, Hawkshaw & Gupta, 2010). A study conducted by Pribuisiene, Uloza and

Saferis (2005) analysed and quantified the voice characteristics of LPR patients to

determine the most important voice tests and voice quality parameters in the functional

diagnosis of LPR. Findings indicated that the mean values of hoarseness on a visual

analogue scale and the total Voice Handicap Index were significantly higher in LPR

patients as compared to asymptomatic patients. A significant increase in mean values of

11

jitter, shimmer and normalised noise energy (NNE) were seen in LPR patients’ acoustic

assessment. Jitter is defined as the variation of fundamental frequency present when

sustaining a vowel, shimmer is the variance in amplitude of vocal folds during sustained

phonation (Colton & Casper, 1996). A significant decrease in pitch range, maximum

frequency, phonetogram area (S) and maximum phonation time (MPT) was seen in the

same LPR patients (Pribuisiene, Uloza & Saferis 2005). Patients with LPR complain

about various combinations of voice symptoms, such as intermittent or chronic

dysphonia, vocal fatigue and voice breaks (Cohen, Bach & Postma, 2002). Generally, the

voices of LPR patients are of poorer quality with restricted phonation abilities compared

to those of healthy people (Jin, Lee, Jeong, Jeong, Lee & Tae, 2008).

Other symptoms frequently seen in LPR, delineated by Sataloff, Hawkshaw and Gupta

(2010), include: morning hoarseness, prolonged voice warm-up time (greater than 20 to

30 minutes), halitosis, excessive phlegm, xerostomia (dry mouth), coated tongue,

possibly geographic tongue, throat tickle, chronic sore throat, dysphagia, aspiration,

regurgitation of gastric contents, nocturnal cough, difficulty breathing (especially at

night), occasionally pneumonia, and laryngospasm. Reimer and Bytzer (2008), and

Schreiber, Garten and Sudhoff, (2009), concur saying that dysphagia and odynophagia

are frequently observed.

Hoarseness is a general term used for any change in the character of a patient’s voice,

and is described by a variety of terms such as breathiness, strain, tremor or roughness

(Ulis & Yanagisawa, 2009). A study conducted by Pribuisiene, Uloza and Jonaitis (2002)

suggested that in cases where LPR is suspected, idiopathic hoarseness as a symptom was

identified in 90, 2% of patients. Other manifestations of the head and neck that have been

reported include otitis media, asthma and sinusitis (Koufman et al., 2002).

Sataloff et al. (2006) also emphasise that LPR may lead to a number of other more

serious maladies, including infant death syndrome, and carcinoma. Subglottic stenosis

and laryngeal cancer have also been associated with LPR (Westcott, Hopkins, Bach,

Postma, Belafsky & Koufman, 2004).

12

A recent case-control study found that patients with reflux were more likely to develop

laryngeal cancer compared to controls (Rees, Pazmany, Gutowska-Owsiak, Inman,

Phillips, Stokes, Johnston, Koufman, Postma, Bailey & Birchall, 2008). Studies suggest

that between 68% and 87% of patients with laryngeal carcinoma have LPR, but caution

that despite this high incidence, a causal relationship can be inferred only after due

consideration of the multifactorial nature of cellular damage and transformation (Galli,

Cammarota, Volante, De Corso, Almadori, & Paludetti, 2006). An important mechanism

highlighted in the development of cancer is chronic inflammation. Alternating regular

tissue damage and repair-phases trigger alteration at a cellular level and ultimately is a

factor that leads to cellular mutation and many tumours. (Galli et al., 2006). LPR is

characterised by the exposure of the larynx and hence laryngeal mucosa to reflux and

may be the root of the chronic inflammation mechanism (Galli, Cammarota, Volante, De

Corso, Almadori, & Paludetti, 2006).

Chronic inflammation is due to the contents of the secretions which are refluxed. These

include hydrochloric acid and pepsin which combine to form chlorhydro-peptic complex.

This complex causes damage at the intercellular junctions of the epithelium leading to

increased permeability and as a result intercellular acidity, alteration of the osmotic

balance and cellular death (Galli et al., 2006). The damage caused by reflux on laryngeal

mucosa combined with additional inflammatory factors (voice abuse, chronic cough,

vomit, recurring infections of the upper airway) would have an influence on the mutation

action of well known cancer-causing extraneous factors such as smoking and alcohol

consumption (Galli et al., 2006).

Laryngeal carcinoma seems to be highly related to the risk factors of tobacco and alcohol.

Awareness is growing that LPR also seems to be an important co-factor, especially in

non-smokers (Postma & Halum, 2006). Koufman (1991) reported that in 31 consecutive

cases of laryngeal carcinoma LPR was evident in 84%, but only 58% were active

smokers. The extent of the relationship between LPR and malignant cellular changes

remains to be fully investigated, but the available data suggest that most patients who

develop laryngeal carcinoma both smoke and have LPR (Koufman, 1991). Additionally,

leukoplakia (a precancerous growth caused by chronic irritation that is visible as white

13

patches) and other premalignant lesions may resolve or significantly improve with

antireflux therapy (Postma & Halum, 2006).

A recent study by Doustmohammadian, Naderpour, Khoshbaten and Doustmohammadian

(2010) examined the association between acid-related inflammation of the upper

digestive tract and laryngeal cancer. Their study involved 65 patients with diagnosed

laryngeal cancer and 65 matched controls. Endoscopic results were collected on all the

patients and results indicated a significant difference between groups. The laryngeal

control group had erosive oesophagitis and gastritis at a statistically higher level than

controls. The researchers concluded that increased acid secretion could account for the

increased severe inflammation and erosion present in the patients with laryngeal cancer.

Further, they concluded that their study supported the hypothesis that gastric acid and

pepsin play a role in laryngeal cancer (Doustmohammadian et al., 2010).

2. 4 Diagnosis of LPR

Diagnosis of LPR is thus important because of its association with laryngeal and upper

aerodigestive tract cancer (Ford, 2005). However, the diagnosis of LPR is controversial

and a number of methods have been postulated to be effective, although, each presents

with a number of shortcomings. The controversy related to diagnosis raises the question

whether LPR may be misdiagnosed, over-diagnosed, or under-diagnosed.

There are 3 principal methods for diagnosis of LPR. Diagnosis can be made on the basis

of ambulatory 24-hour double probe pH monitoring, impedance testing, as well as

symptoms and laryngeal findings (Koufman, et al., 2002). However, lack of consensus on

diagnosis of LPR has arisen from issues with all of these methods.

pH testing is considered the diagnostic gold standard for LPR by some (Habermann,

Kiesler, Eherer, 2002; Barry & Vaezi, 2010). This is because reflux laryngitis is deemed

to be caused by episodes of acid and peptic reflux into the laryngopharynx (Joniau et al.,

2007). Dual pH probes are conducted over an 18 to 24 hour period in an outpatient

setting. The lower probe is placed approximately 5 centimetres above the lower

oesophageal sphincter. The upper probe is placed just above the upper oesophageal

14

sphincter. Placement is made with the assistance of manometry or endoscopy (Postma &

Halum, 2006). The upper oesophageal sphincter is seen as the final barrier against LPR.

Therefore placement of the probe at this point ensures that if measurements are made at

the upper probe and indicate the presence of LPR, is in all likelihood taking place.

There is however, controversy regarding the efficacy pH testing in LPR diagnosis,

(Vaezi, Hicks & Abelson, 2003) be it with single or dual probes (Abaza, 2004). False

negatives may result from the intermittent nature of LPR or pH may be neutral possibly

due to suggested changes in pepsin and non-acid content of the refluxate as suggested by

Abaza (2004).

LPR signs can occur in patients with weakly acid or non-acid reflux. Reflux is classified

as either being acidic or non-acidic based on the pH of the refluxate measured by the

probes during monitoring (Mainie, Tutuian, Shay, Vela, Zhang, Sifrim & Castell, 2006).

Gastric refluxate typically contains both pepsin and acid. Pepsin is an enzyme produced

in the stomach as pepsinogen. However, when pepsinogen is exposed to acid it is

converted to pepsin and is able to digest protein (Printza, Speletas, Triaridis & Wilson

2007). Therefore when pepsin is absorbed by laryngeal cells and reactivated by a drop in

pH as seen in LPR, this causes further damage to the larynx (Gupta & Sataloff, 2009).

This mechanism provides one possible explanation for the finding of LPR symptoms and

signs in patients with weak or non-acid reflux (Gupta & Sataloff, 2009).

Patients with nonerosive reflux disease (NERD) have less lower oesophageal acid

exposure than patients with Erosive Oesophagitis (EE). A study by (Savarino, Tutuian,

Zentilin, et al., 2010) as cited by Lee, Kim, Ryu, Kim, Cheong, Lee and Song, (2010)

compared reflux episodes and symptoms associated in NERD and EE patients using

combined Multichannel Intraluminal Impedence (MII)/pH-metry. They reported that acid

reflux episodes, volume, and acid clearance were crucial factors in the cause of reflux

induced lesions (Lee, Kim, Ryu, Kim, Cheong, Lee & Song, 2010).

Joniau et al. (2007) reviewed 13 double pH probe studies on patients diagnosed with LPR

and 11 studies on normal controls. The findings reveal a low number of reflux events in

diagnosed LPR patients and only a marginally higher incidence than normal controls. No

15

statistical differences were found between the prevalence of pharyngeal reflux events in

normal controls or diagnosed patients (Joniau et al., 2007).

Disagreement abounds regarding the normative data and values related to reflux in pH

studies (Ford, 2005). Normal pH values for the hypropharynx are not well established

(Printza, Speletas, Triaridis & Wilson, 2007). However, the aim of a recent study was to

use a new probe measure in order to try to establish the pH thresholds in normal subjects

in order to be able to identify abnormality in pH (Ayazi, Lipham, Hagen, Tang, Zehetner,

Leers, Oezcelik, Abate, Banki, & DeMeester, 2010).Results from this study indicate that

even normal subjects have distinctly different pH thresholds in upright, (pH of 5.5),

compared to supine ,(pH of 5.0), positions (Ayazi et al., 2010).

Smaller changes in pH are necessary to damage laryngopharyngeal tissue, than for

oesophageal tissue (Ford, 2005), although lower oesophageal standards for GORD are

often applied to LPR diagnosis. Ford (2005) suggests that a drop in pH to less than 5 may

better indicate LPR as neutralising factors such as saliva and airway secretions can raise

pH values. In addition, the potential harm from non-acid or gaseous refluxate is not

accounted for in pH tests (Ford, 2005). This system may be susceptible to damage by

non-acid reflux due to mucosal differences of the oesophagus and the upper aerodigestive

tract.

Furthermore, a study by Johnston et al. (2004) suggests that exposure to pepsin

during LPR causes a decrease in carbonic anhydrase III in the laryngeal epithelium

which results in reflux related inflammatory laryngeal injury. Carbonic anhydrase

III is responsible for producing bicarbonate ions that neutralise refluxed gastric acid

by entering the extracellular spaces. Therefore a reduction in carbonic anhydrase

would translate into less protection of mucosa due to decreased bicarbonates,

enabling the pepsin to remain active and cause cellular damage.

Smaller pH changes in gaseous refluxate, a decrease of more than 1 (>1), have been

found increasingly in patients with LPR compared to those with GORD or healthy

controls (Ford, 2005). Joniau et al. (2007) postulate that unreliability in pH tests leads to

significant over-diagnosis of this disease.

16

A study by Muderris, Gokcan and Yorulmaz (2009) indicates that placement of the pH

probes during measurement is crucial to accurate identification of true hypopharyngeal

reflux episodes. They suggest that a bifurcated, triple-sensor pH probe allows for

monitoring of the oesophagus and hypopharynx simultaneously. Ideal placement of the

probes is above the upper oesphogeal sphincter, one below the upper oesophageal

sphincter and one in the distal oesophagus. Single probe, double sensor pH monitoring is

more commonly utilised in LPR diagnosis and Muderris et al. (2009) suggest that greater

accuracy is achieved if the proximal probe is placed in the pharynx as opposed to the

upper oesophagus. Clearly placement of the probe will have an impact on whether

occurrences of refluxate are accurately monitored and recorded. This in itself poses a

problem with diagnostic accuracy as probes may shift or reflux may occur at instances

when the patient is not being tested (Gupta & Sataloff, 2009).

Further evidence against pH monitoring is found in a study where patients with

pharyngeal reflux documented by pH monitoring were no more likely to respond to acid-

suppressive therapy than patients with no documented reflux (Ulualp, Toohill & Shaker,

2001). These findings along with the lack of consensus regarding placement of the pH

probe, decreased consistency in the pH being measured and the intermittent nature of

LPR reduces the enthusiasm for pharyngeal pH monitoring in the diagnosis of LPR.

Impedance testing is a more recent form of diagnosis and relies on a catheter that

measures electrical resistance (impedance) between different points along the

oesophagus. It is superior to pH monitoring in that it is able to detect the reflux of acid

and non-acid liquid or gaseous material. (Barry & Vaezi, 2010). Liquid boluses decrease

impedance while gaseous boluses increase the impedance between electrodes (Maron &

Jordaan, 2010).

Another method that seems to have had some success in diagnosing LPR is the

ambulatory multichannel intraluminal impedance (MCII) (Srinivasan, Vela, Katz,

Tutuian & Castell, 2001). MCII functions by measuring changes in the resistance of

intraluminal gases, liquid, or bolus to alternating current between a series of metal

electrodes, placed in the oesophagus (Srinivasan et al., 2001). MCII is able to detect the

direction of the flow of small amounts of material in the oesophagus (Abaza, 2004). This

17

is important in those patients who, experience non-acid or weak acid reflux as refluxate

can still be detected regardless of the pH (Abaza, 2004).

Other patients who may not be suitable candidates for regular dual probe pH monitoring

can also be successfully identified using this method. Another advantage is that the pH

level, as well as the normative standards for the oesophagus and the larynx become less

important, provided that it is was clear that refluxate is reaching the laryngopharynx and

therefore causing damage. This method of diagnosis is often reserved for those whom, do

not respond to empirical treatment regimes, thus warranting alternative diagnostic

investigation (Bove & Rosen, 2006). Because non-acid reflux events in symptomatic

patients who have had negative pH probe studies may therefore be detected (Postma &

Halum, 2006), Musser et al. (2009) have suggested that impedance testing may provide

greater agreement amongst diagnostic methods and ultimately diagnosis of LPR may

become less elusive.

Due to the absence of definitive diagnostic criteria laryngopharyngeal reflux remains a

subjective entity according to, Bove and Rosen (2006). Attempts to increase the

reliability of clinical investigation for identifying patients with LPR led to the

establishment of the Reflux Severity Index (RSI) and the Reflux Finding Score (RFS)

(Joniau et al., 2007). The development of such scales has begun to remedy the diagnostic

void (Bove & Rosen, 2006). pH testing and impedance measures of assessment are then

only recommended for those patients who do not respond to treatment (Bove & Rosen,

2006).

The Reflux Finding Scale (RFS) (Appendix C) is a well known multipoint scale designed

by Belafsky, Postma and Koufman (2001b) that rates clinical signs and appearance of the

larynx and pharynx in order to determine the presence and severity of LPR. The scale

investigates eight laryngoscopic signs: subglottic oedema, (swelling below the glottic

region), ventricular obliteration, (refers to oedema of the true and false vocal folds

appearing to obliterate or obscure visualisation of the laryngeal ventricle), erythemia or

hypermia (erythemia of the arytenoid cartilages), vocal cord oedema (oedema of the false

18

and true vocal folds), diffuse laryngeal oedema, posterior commissure hypertrophy

(mucosa of the posterior commisure is hyperatrophied and becomes swollen to varying

degrees), granuloma or granulation tissue (most commonly in the region of the vocal

process) and the presence of thickened endolaryngeal mucus (Abaza, 2004; Barry &

Vaezi, 2010; Sataloff, Hawkshaw & Gupta, 2010; Postma & Halum, 2006; Belafsky,

Postma et al., 2002).

The total reflux finding score can range from 0 (best) to 26 (worst) (Barry & Vaezi,

2010). The authors of the scale determined that an individual with a score higher than 7

can with 95% certainty be diagnosed as having LPR (Belafsky et al., 2001b).

There have been conflicting reports on the validity and reliability of the RFS. Koufman et

al. (2004) indicate that the RFS has been shown to be a valid scale with good inter- and

intra -rater reliability. Joniau et al. (2007) contradict this in their paper highlighting that

only laryngeal granulations have been associated with reflux and claim that validation of

the RFS has been weak. In contrast, a study by Aviv, Liu, Parides, Kaplan and Close

(2000) indicated that posterior laryngeal oedema is the hallmark of LPR. Joniau et al.

(2007) noted that laryngeal findings commonly ascribed to LPR can be identified in as

many as 64%- 86% of normal controls. This indicates that laryngoscopic findings have

decreased specificity or sensitivity to LPR diagnosis. A study by Branski, Bhattacharyya

and Shapiro (2002) cited by Barry and Vaezi (2010) indicate the diagnosis of LPR tends

to be subjective as laryngeal findings are often nonspecific. Barry and Vaezi (2010) feel

that the RFS as a diagnostic method has not been validated in a large-scale randomised

trial and consequently has yet to be incorporated into routine otolaryngology practice.

Despite this, laryngeal clinical findings as a means of diagnosis are widely utilised.

Patient report and rating of symptoms is another method commonly used in clinical

practice. The Reflux Severity Index (RSI) (Appendix B) was developed by Belafsky,

Postma & Koufman (2002). They advocate the use of this scale in conjunction with the

RFS. The RSI is a 9 item self-rating system scaled from 0-5, in which the patient rates

reflux symptom frequency (Koufman et al., 2002). This subjective scale is utilised in

measuring changes beyond clinical laryngeal findings (Abaza, 2004). The maximal score

19

is 45 points and a score of greater than 5 strongly suggests LPR (Belafsky, Postma, Amin

& Koufman, 2002). The symptoms outlined in the RSI include: hoarseness, throat

clearing, excessive mucus or postnasal drip, difficulty swallowing, coughing, breathing

difficulties, a sensation of a lump in the throat, heartburn, chest pain or indigestion

(Belafsky, 2008).

The symptoms described above are not however, exclusive to reflux related damage, and

may be associated with smoking, voice abuse, allergies, viral infections, inhaled

environmental irritants and alcohol abuse (Ford, 2005; Reimer & Bytzer, 2008). Diseases

that have similar symptoms may thus be mistakenly labelled as LPR. Belafsky, Rees,

Rodriguez, Pryor and Katz (2008) suggest that some signs and symptoms of LPR may be

attributed to oesophopharyngeal reflux (EPR) and LPR may therefore be over diagnosed.

A combination of three symptoms, hoarseness, throat itching, and globus pharyngeus

were found to separate LPR patients from healthy individuals significantly (Pribuisiene et

al., 2002). However, a recent study on the RSI questionnaire indicated that it did not

adequately capture the full range of potential reflux symptoms regularly encountered in

otolaryngology patients. This study may in fact contribute to the ongoing uncertainty

about the role of acid or pepsin suppression (Papakonstantinou, Leslie, Gray, Chadwick,

Hudson, & Wilson, 2009). If this is indeed the case then a more comprehensive scale

may be needed, although independent Israeli and Italian studies both found good

reliability and clinical validity of the RSI in identifying patients with LPR when it was

translated into Hebrew and Italian (Schindler, Mozzanica, Ginocchio, Peri, Bottero &

Ottaviani, 2010; Cohen, Gil & Fliss, 2005)

The presentation of Paradoxical Vocal Fold Movement (PVFM) disorder may mimic a

multitude of laryngeal and pulmonary entities such as asthma, LPR disease and polyps

amongst others as well as dystonia and laryngospasm (Murry, Tabaee & Aviv, 2004).

Rosen et al., (2004) claim that patients with LPR and noncompliant cough in the absence

of pulmonary disease should be evaluated for PVFM disorder. This is due to the

symptoms of PVFM disorder being both laryngeal and respiratory in nature including:

wheeze, cough, dyspnea, choking sensation, chest pain, stridor, reflux, dysphonia,

dysphagia (Murry et al., 2004). A number of these symptoms are used for differential

20

diagnosis of LPR, and therefore PVFM disorder may be co-occurring with LPR or LPR

indeed may be mistaken for PVFM disorder.

Various other methods of diagnosis have been suggested. One such method is a trial

treatment regime (Pribuisiene, Uloza & Jonaitis, 2002). Diagnosis in this method is

usually based on the response of LPR symptoms to empirical treatment (Bove & Rosen,

2006) with proton pump inhibitors (PPI’s). The success of this method will be discussed

later under management of LPR. The success of the diagnosis then seems to depend on

the success of the treatment. Therefore if treatment fails was the patient incorrectly

diagnosed or is treatment not sufficient for other reasons?

This lack of specificity in all of the prevalent methods of diagnosis, may lead to a degree

of over-diagnosis of LPR disease (Joniau et al., 2007, Randhawa, Mansuri & Rubin,

2010), or alternatively under-diagnosis, leading to prolonged symptoms, delayed healing

and decreased quality of life (Ford, 1999). Both can lead to unnecessary costs (Ford,

1999).

Furthermore, a distinct correlation between symptoms and laryngitis is unclear (Reimer &

Bytzer, 2008) A recent study by Musser, Kelchner, Neils-Strunjas and Montrose (2009)

found that LPR symptoms and physical findings of LPR were not significantly associated

on the RSI and the RFS. In cases of LPR the laryngeal findings are not always associated

with symptom severity (Khan, Hashmi, Elahi, Tariq & Ingrams, 2006). A study by

Qadeer, Swoger, Milstein, Hicks, Ponsky, Richter, Abelson and Vaezi (2005) found that

the most common symptoms in LPR were sore throat (40%), hoarseness (30%), and

cough (20%), whereas the most common signs were medial arytenoid wall

erythemia/oedema (60%), interarytenoid erythemia (50%), and arytenoid complex

erythemia/oedema (50%). The study involved 72 patients with symptoms of LPR and

laryngoscopic findings suggestive of LPR despite aggressive acid suppression therapy.

Each patient underwent fundoplication, a surgical treatment for reflux, and was

monitored post operatively for a year using a daily record of symptoms and laryngeal

signs. Patients were assessed at 6 and 12 months post-fundoplication via laryngoscopy.

Interestingly laryngeal signs resolved in 80% of patients by the 12 month laryngoscopy

indicating successful surgery. However, only 10% of patients had symptomatic

improvement. Therefore neither signs nor symptoms may be used as a substitute for the

21

other. The study concluded that there appears to be poor correlation between signs and

symptoms of LPR.

This study confirmed earlier findings by Belafsky, Postma and Koufman (2001a) in

which laryngeal signs of LPR continued to improve while symptoms of LPR only showed

limited improvement following a 6 month treatment with twice-daily PPI’s.

A survey of approximately 2,000 gastroenterologists and ENT physicians indicated that

oedema and erythemia of the larynx were the most common signs used for LPR diagnosis

(Ahmed, Khandwala, Abelson, Hicks, Richter, Milstein, & Vaezi, 2006). This however,

should be done with caution as studies have found that 86% of healthy volunteers had at

least one sign attributed to LPR (Hicks, Ours, Abelson, Vaezi, & Richter, 2002; Milstein,

Charbel, Hicks, Abelson, Reicheter, & Vaezi, 2005). Specifically 70% of the normal

controls had interarytenoid bar, 29% had erythemia on the medial edge of the arytenoids

and 21% had posterior pharyngeal wall cobble stoning (Hicks et al., 2002). Alarmingly

they also found that different instruments used for endoscopic evaluation lead to different

laryngoscopic findings. Flexible endoscopy was able to detect abnormalities in 93% of

the population compared with 83% through the rigid endoscope (Milstein, Charbel,

Hicks, et al., 2005).The study by Qadeer et al. (2005) suggests a multifactorial aetiology

for LPR, since most of the authors patients’ symptoms improved only after adjunctive

therapy for other conditions.

2. 5 Management of LPR

Difficulties experienced with effectively treating LPR raise questions regarding accuracy

of diagnosis. Three conventional methods of treatment are outlined in the literature (Ford,

2005). These include; lifestyle and dietary changes; medication and surgery.

Koufman et al. (2002), suggest that mild and/or intermittent cases of LPR should be

treated with dietary and lifestyle modifications, as well as with an H2 antagonist.

Lifestyle and dietary changes recommended are limiting fat intake, cessation of smoking,

avoidance of caffeine, carbonated drinks, alcohol and acidic fruit juices. Further, not

eating within 3 hours of bedtime; and limiting wearing tightly fitted clothing around the

waist (Sataloff, 2005). Some research indicates that lifestyle and dietary changes are less

22

effective in patients with LPR than those with GORD (Abaza, 2004). Bove and Rosen

(2006) state that LPR is seldom managed with behaviour modification alone.

Medical management is the most commonly utilised treatment method (Abaza, 2004).

Proton Pump Inhibitors (PPI’s) are the mainstay of therapy in LPR, although they are not

without controversy regarding their efficacy (Reimer & Bytzer, 2008). PPI’s are the drug

of choice in GORD and are highly effective in resolution of long-term acid-related

symptoms (Reimer & Bytzer, 2008). However, management of LPR using PPI’s are

successful in as few as 60% of patients (Belfasky et al., 2008). This has caused some

clinicians to question the role of gastric refluxate in these patients (Belfasky et al., 2008).

Difficulties with PPI management may be attributed in some instances to contents of the

proximal oesophagus being refluxed into the laryngopharynx, and so it has not had time

to mingle with gastric juices. The American Academy’s Speech and Swallowing

Committee recommended a 6 month empiric treatment with bid PPI for LPR as initial

treatment (Abaza, 2004). More prolonged and aggressive treatment for LPR compared to

GORD is needed according to, Koufman et al. (2002). However, a resistance to PPI’s in

those who, initially respond to treatment may develop (Sataloff et al., 2006). A study by

Karkos and Wilson (2006) could not establish a significant effect of this medication over

placebo on the symptoms and signs of reflux laryngitis (Joniau et al., 2007).

The failure of a percentage of patients to successfully respond to PPI therapy may be due

to the fact that the laryngeal and pharyngeal symptoms associated with reflux are very

common and can also be due to a number of other aetiologies such as -: allergy, smoking,

asthma, infections, voice abuse or alcohol abuse (Printza, Speletas, Triaridis & Wilson,

2007). Specifically a number of patients with allergic laryngitis are misdiagnosed with

LPR and therefore fail treatment with proton pump inhibitors (Randhawa, Mansuri &

Rubin, 2010). The diagnosis of such cases may be called into question. PPI therapy as a

means of management seems to be inaccurate and erratic as symptoms (as measured by

the RSI), laryngoscopic findings (as measured by the RFS), or abnormal findings on pH

monitoring will not predict response to PPI therapy (Reimer & Bytzer, 2008). In

juxtaposition to this Hammer (2009) indicates that a reflux symptom index and a reflux

finding score may be beneficial in assisting with selection of the minority of patients

who, may profit from acid-suppressive therapy.

23

Voice problems are associated with LPR and require a team approach (Sataloff, 2005b).

Care is provided by the otolaryngologist, speech therapist and voice and singing coach,

when applicable. The otoloaryngologist plays a key role in diagnosis and medical

management of LPR. The voice and singing coach teach the patient to use various vocal

and body movement techniques to optimise physical function and break down tension

(Sataloff, 2005b). The singing coach promotes relaxation, abdominal and thoratic muscle

strength as well as breath control (Sataloff, 2005b). Chronic throat clearing and a feeling

of a lump in the throat, symptoms of LPR according to, Belafsky et al. (2008), may lead

to vocal abuse and consequently may cause vocal chord oedema, contact ulcers and

polyps (Martin, 1987). The speech therapist will attempt to improve the vocal symptoms

related to these physiological and anatomical changes through various techniques.

However, it is likely that a combination of treatment options is able to give the best and

most sustainable results in treating LPR.

For those who do not experience any relief with the individual or combined approaches to

treatment there is surgical management. Surgical intervention is reserved for severe cases

of LPR, non-acid reflux and those who have not found success with alternative

management (Bove & Rosen, 2006). Laprascopic Nissen fundoplication surgery is found

to be useful in certain patients (Abaza, 2004; Lindstrom, Wallace, Loehrl, Merati &

Toohill, 2002). Restoration of lower oesophageal competence is the goal of surgery

(Ford, 2005). Positive feedback has been reported in a study by Oeschlager, Eubanks,

Oleynikov, Pope and Pellegrini (2002) where a significant decrease in pharyngeal reflux

was reported in patients who, underwent surgical intervention.

It must however, be noted that one study of surgical fundoplication indicated that surgery

is not recommended for patients whose symptoms do not respond to aggressive PPI

therapy (So, Zeitels & Rattner, 1998). Alternatively Remacle and Lawson (2006) suggest

that surgical intervention is a good option for patients with observable high volume reflux

and abnormal lower sphincter competence when medical management fails. Despite these

conflicting reports, surgical intervention must be considered as a last resort for those

patients with severe LPR, and it may not be successful in resolving or even reducing

LPR. However, for some, nissen fundoplication’s role in LPR has yet to be defined

(Westcott, Hopkins, Bach, Postma, Belafsky & Koufman, 2004).

24

Treatment of LPR is difficult and not consistent across all patients. What works in some

cases has no impact on others. Treatment is thus arduous and controversial due to

diagnosis being vague and challenging (Joniau et al., 2008).

‘Because of a paucity of convincing evidence regarding techniques for

establishing definitive diagnosis and causation in individual patients, and

because of a plethora of imperfect studies that have produced conflicting

conclusions, LPR diagnosis and management remain controversial’

(Sataloff, Hawkshaw & Gupa, 2010 p.214).

This research therefore aimed to establish the relationship between the Reflux Severity

Index (RSI) and the Reflux Finding Score (RFS) in participants that have attended the

Wits University Donald Gordon Voice and Swallowing clinic. The aims were:

to establish if there is a correlation between the total RFS and RSI scores

to ascertain if certain test items of the RSI and RFS are elevated in said patients

to describe trends related to RFS and RSI sub-scores and

to verify if extraneous factors such as age, gender, professional voice use and

smoking impact on the sub scores of the RFS and RSI and to describe trends

based on these variables.

25

Method 3.1 Aims

The primary aim of this research was to establish the relationship between ratings on the

RSI and RFS in participants who attended the WITS University Donald Gordon Voice

and Swallowing clinic. The subordinate aims were:

(1) to establish if there was a correlation between the total RFS and RSI scores

(2) to ascertain which test items of the RSI and the RFS were elevated in the participants

(3) to describe trends in RFS and RSI sub scores and

(4) to determine if extraneous factors such as age, gender, professional voice use and

smoking impact on the sub-scores of the RFS and RSI and to describe the trends based on

these variables.

3.2 Research Design

The research study employed a retrospective descriptive quantitative design with a

correlational within-subject component and a comparative between-subject component

(Schiavetti & Metz, 2002).

Retrospective research enables investigation of a large amount of data that has been

collected prior to the formulation of the research question (Schiavetti & Metz, 2002).

This is particularly relevant to this study as 105 patient’s charts were reviewed. The

charts contained information that had been previously collected between 2005 and 2008

at the clinic where the records are stored. This aspect of the design is crucial as it allows

the data which may not have been collected with the formulated research question to be

used for research purposes.

A descriptive design is used to observe differences in a group, developmental trends, or

interaction among variables that can be measured by the researcher (Schiavetti & Metz,

2002). In this study the RSI and the RFS results were evaluated. Trends and interactions

among the different subscores were described.

26

Quantitative data is also known as measurement data and refers to one of the two kinds of

numerical data. Quantitative data is obtained by measuring objects or events (Howell,

1999). In this study the scores on the RSI and the RFS were measured as well as the

scores on the individual items for a given sample of participants.

Correlational research is an example of a within-subject design in descriptive research as

a number of measures are applied to a group of participants (Schiavetti & Metz, 2002).

An advantage of a within-subject design is that the problem of extraneous variables

impacting on one group of participants and not the other is eliminated (Shiavetti & Metz,

2002). Within the current research all the participants were assessed on the RFS and the

RSI, and the relationship between the two measures was evaluated. A correlational

research design is employed to study the relationship among two or more variables by

evaluating the degree to which changes in one variable affects change in the other

variable (Schiavetti & Metz, 2002). One of the major components of the current research

project aimed at identifying the relationship between variables on the RFS and RSI and

gaining insight into the impact that change in one variable has on another variable.

The comparative component involved comparing groups of participants in terms of age,

gender, smoking status and professional voice use to evaluate the effects of these

variables on the RFS and RSI ratings.

3.3 Participants

Inclusion Criteria

The participants were selected based on the following criteria.

Firstly, they had to have attended the Donald Gordon Voice and Swallowing Clinic

(DGVS). All of the patients attending this clinic will have a voice or swallowing disorder

that may be reflux related, as discussed in the literature review. In the initial stages of

data collection, patients files from the DGVS clinic were utilised. During the data

collection phase the files were moved to the Netcare Parklane Clinic, where the

remainder of the data was collected. This did not impinge on the study as all of the

patients RFS and RSI scores had been completed whilst they were attending the DGVS

clinic. Therefore, no interference from the new site was evident due to the retrospective

27

nature of the research. A clinic similar to that held at DGVS clinic is conducted at

Netcare Parklane Clinic. The Netcare Parklane clinic is serviced by the same team of

specialists that scored and evaluated the scales for each patient in this study.

Secondly, each participant must have completed the Reflux Symptom Index questionnaire

themselves prior to consultation with the multidisciplinary team. Only patients with a

total RSI score of greater than 5 (>5) were included in the study as according to,

Belafsky, Postma, Koufman (2001b).

Thirdly, patients must have undergone both a trans-nasal and trans-oral fibreoptic video -

stroboscopy so that their laryngeal status had been observed and recorded. This allowed

the researcher to obtain the Reflux Finding Score.

Rigid and flexible endoscopy was conducted with each patient for a number of reasons

including:

“Videostroboscopy is performed with a rigid endoscope placed in the mouth

up to the tongue base or a flexible fibre-optic trans-nasal endoscope with a

xenon light source passed through the hypo-pharynx for an unimpaired view

of the larynx. Flexible fibre-optic trans-nasal endoscopy is preferred for

evaluation of neurolaryngeal disorders (including hypokinetic dysarthria,

spasmodic dysphonia, essential tremor, and paradoxical vocal fold motion) for

best visualisation of vocal fold mobility, differentiation of vocal fold paralysis

from paresis, hypo- and hyper-functional vocal fold behaviours, and laryngeal

muscle tone and agility. The latter is also used when examination with a rigid

endoscope does not allow adequate visualisation due to structural interference

or patient discomfort.”

(Thibeault & Zelazny, 2010, paragraph 45)

28

The two approaches are thus used for different purposes and are important in differential

diagnosis. Previous research has shown that flexible endoscopes significantly

underrepresented reflux signs and a rigid laryngeal endoscope should be utilised as an

ideal when examining the larynx for signs of LPR (Eller, Ginsberg, Lourie, Heman-

Ackah, Lyons & Sataloff, 2009). However, contradictory findings indicate that signs such

as interarytenoid bar, erythemia of the medial wall of the arytenoids, considered to be

signs in LPR, were more often detected with flexible than with rigid laryngoscopes,

suggesting that flexible laryngoscopy is more sensitive in identification of laryngeal

tissue irritation (Milstein, Charbel, Hicks, Abelson, Reicheter, & Vaezi, 2005). Sataloff,

Hawkshaw & Gupta (2010), also indicate that flexible laryngoscopy may be useful in

patients with a hyperactive gag reflex which is common in patients with severe LPR.

Therefore there are pros and cons for each of these endoscopes and both were utilised as

they provide the most holistic picture of the larynx.

The majority of patients included in the study were adults who are more reliable in terms

of their completion of the RSI which is a subjective rating scale. Children may have

difficulty understanding what the questionnaire is asking them, and parents may not be

able to adequately record the child’s experience of a symptom as they do not experience

the symptom themselves.

Further, children are assessed using a number of different rating scales compared to those

utilised with adult patients. This is because many of the symptoms for LPR must in fact

be visible or noticeable to an external person (such as a parent) in order for them to be

scored on a scale. Adult’s symptoms of LPR may be subjective and only noticed by the

patient for example lump in the throat. The children’s scales were developed to account

for these factors and have been tested for reliability and validity. Three well known

surveys of voice for adults have been adapted to the paediatric population. Firstly, the

Voice Outcome Survey has been adapted to create the Paediatric Voice Outcome Survey

which is a four-item parental proxy.

Secondly, the well utilised voice related Quality of Life questionnaire adapted to a 10

item parental substitute, the Paediatric Voice Related Quality of life questionnaire.

Thirdly, the voice Handicap index which was adapted to a 23 item parent replacement –

Paediatric Voice Handicap Index (Verduyckt, Remacle, Jamart, Benderitter, &

Morsomme, 2009). The youngest participant included in the study was 12 years old.

29

Although this participant is legally not an adult, the literacy abilities and understanding of

a 12 year old was considered adequate. This implies that the participant would be able to

read the RSI independently or have it read to him, answer the scale questions accurately

and therefore be included in the study.

Sampling

A retrospective chart review of 250 patients seen at the DGVS Clinic from 2005 to 2008

was conducted. Of these files, 105 records were considered viable and were included in

the research, based on the selection criteria described above.

Sample Recruitment

Participants were recruited through the use of written informed consent during their first

consultation at the DGVS clinic. Each patient attending the clinic, decided prior to

consultation if their files and data may be used for research purposes (Appendix A). If

patients met the inclusion criteria and had provided written consent for the use of their

records they were included in the study. The flow diagram in figure 3.1 represents the

process of sample recruitment.

30

Diagram 3.1 Flow Diagram representing the procedure for Data collection

Description of Patients

One hundred and five (n=105) patients’ records were included in the study. The

participants are described in Table 3.1.

Patient attends Voice and Swallowing

clinic

Patient reads request for written informed

consent for use of records for research

Patient provides consent

Patient does not provide consent

Patient completes the RSI. Team assesses

patient on video stroboscopy

Patient completes the RSI. Team assesses

patient on video stroboscopy

Patient data is used in the current LPR

research

Patient data is filed in personal file at the clinic and not used

for research purposes

31

Table 3. 1. Description of patients

Total

patients Male Female Smoker

Non-

smoker

Prof. voice

user

Non-prof.

voice user

105 43 62 21 84 50 55

Of the sample 43 participants (40,95%) were men and 62 participants (59,05%) were

women. This does not necessarily translate into females being more prone to LPR than

males. This difference in gender within the sample is due to the fact that the sample was

not age and gender matched. Rather participants were incorporated into the study based

solely on the inclusion criteria and not extraneous variables. It may in fact be that women

attend to health related queries before men do, or that females feel that a vocal problem is

more of an issue than men do based on their background and propensity for

communication and speaking. Alternatively there may be more LPR in females than

males based on minute structural differences in enzymes, histology and general levels of

sensitivity. We cannot make these assumptions based on the information that we have on

this topic. This may be a topic for further research.

The age range of the sample was from 12 years to 89 years with highest proportion of

patients (n=21 or 20% ) , (n=19 or 18%) for patients in the age range from 20–to-29 years

and 30-to-39 years of age respectively. The 50-59 year old age range comprised 17 (16%)

participants while the 40-49 year decade comprised 16 (15%) participants and the 60-69

year old category (13%; n=14) of the sample. The lowest age range from 12 years to 19

years comprised (8%; n=9) while the upper end of the age spectrum, 70-79 years

reflected (5%; n=6) of the sample. The smallest number of participants were in the 80-89

year old population with only 3 (2%) of the sample in this range. This distribution may be

affected by a number of factors and again a relationship between age and LPR prevalence

can not be inferred on this data alone. However, age and its relationship with the signs of

LPR as evaluated on the RFS and symptoms as evaluated by the RSI were explored

statistically and descriptively as part of this research report.

32

The mean age of participants was 43,4 years. The median age was 41 years. (The median

refers to the number in the range that is in the centre or middle of the data set.) The

sample demonstrated a bimodal distribution with 27 and 61 years of age representing the

modes. (Mode refers to the highest or most commonly occurring number within the data

set). The wide age range contributes to the external validity of the study, as these results

have applicability to a greater age range of the general population. Table 3.2 shows the

mean, mode and median of the ages of patients included in the study.

Table 3. 2. Measures of central tendency of patients’ age

Mean age Median Mode

43,4 years 41 years 27 & 61 years

Within the sample, 21 (20%) of the participants were smokers and 84 (80%) were non-

smokers. This represents an unexpectedly large difference between the number of

smokers and non-smokers in the research group. This difference may be attributed to

social factors, such as socio economic background, or perhaps fewer people smoke due to

widespread knowledge regarding the health risks of smoking. This may also be linked to

the higher proportion of females in the sample compared to males. Young males are more

likely to be smokers than middle aged females. The mode of the sample was 61 years and

this population may have reduced smoking tendencies due to other health issues they may

experience e.g. diabetes etc. However, perhaps the most obvious explanation is that

patients who, suffer from LPR had noticed a reduction in respiratory difficulties and

cough if they reduce or cease smoking. Smoking status as a variable impacting on the RSI

and RFS ratings was also examined.

Of the sample 50 (47,6%) were professional voice users. This is marginally lower than

non professional voice users. Despite this small difference it confirms the notion that this

group does constitute a significant proportion of the case load at a voice clinic. This

population does seem to have a propensity for developing reflux. Professional voice users

33

are justifiably even more anxious about their voices than the rest of the population

because a voice problem is a threat to their livelihood.

These two subgroups have a small amount of overlap as can be seen from diagram 3.2

below. Of the 105 participants,46,7% (n=49), were non-smokers and non-professional

voice users. Non-smoking professional voice users made up 33,4% (n=35) of the sample.

Professional voice users who, smoked made up 14,2% (n=15), of the sample and smokers

who, were non-professional voice users constituted only 5,7% (n= 6) of the sample.

Therefore 30% of the professional voice users in the sample were smokers compared to

11% of non-professional voice users being smokers. More than a quarter of the

professional voice users are smokers. This is surprising as one would expect professional

voice users to be aware of the effects that smoking may have on respiration, voice and

damage to the oral and laryngeal structures. It is commonly recommended that

professional voice users should take these risks more seriously as their careers depend on

the optimal functioning of their voice. It is clear that the majority of patients in the

sample are non-professional voice users who, do not smoke.

49

35

15

6 nonsmoker non prof

voice user

nonsmoker prof voice

user

smoker prof voice user

smoker non prof voice

user

Diagram 3. 2 Pie chart illustrating the relationship between smoking status and

professional voice status in the sample (n=105).

34

The sample size was large enough to allow for results to be generalised with more

confidence and leads to better external validity of the study (Schiavetti & Metz, 2002).

The large sample size also allowed for a representative sample of each category of

participants in terms of gender, age and professional vs non-professional voice users.

3 4 Ethical considerations

This research hinged on patients records. Patients are deemed to be at risk populations

and research using their records necessitates ethical clearance from the University of the

Witwatersrand Medical Ethics Committee. Ethical clearance was obtained in April 2009

Protocol number: M090467 (Appendix E).

Ethical considerations necessitated obtaining written informed consent from patients of

the WITS University Donald Gordon Voice and Swallowing Clinic to have their records

and results utilised for research purposes. On the initial visit to the WITS University

Donald Gordon Voice and Swallowing Clinic each patient was given a written consent

form (Appendix A). The form was explained and each patient asked if they would

consent to have their files and data utilised for research purposes. The patient is not under

any obligation to agree and there are no negative consequences to indicating they would

not like their results to be used for research purposes. This was explained to each patient

and they were also told that they were entitled to withdraw their consent at any point

without any implications. The patient is requested to complete the informed consent form

marking either that they do consent to their details being utilised for research or they do

not consent. This written informed consent form is filed in their patient files at the Donald

Gordon Voice and Swallowing clinic.

Patients were reassured that personal information will not be utilised or documented in

the research. In this study, a coding system, assigning a number to differentiate patients’

ratings, assured that the research was anonymous and confidential. The retrospective

methodology also protected the patient, taking into account their rights as well as

enabling the researcher to extract and examine information in a way least intrusive to the

35

patient. The researcher has accounted for reliability and validity threats that may interfere

with data collection and analysis.

3.5 Measures

Patient Rating Scale:

Symptoms of laryngopharyngeal reflux were measured using the Reflux Symptom Index,

a scale developed by Belafsky et al. (2002). The self-rating scale allows 9 items to be

rated from 0-5 by the patient. This scale is typically used in otolaryngology practices to

establish the subjective perception of possible reflux (Appendix B). Patients are required

to fill in the RSI prior to seeing the team of specialists at the clinic. Patients are instructed

to fill in the scale as best as they can, rating each symptom on the scale from 0 = no

problem to 5 = severe problem. The patient is requested to rate each of the items on the

scale according to, how the symptoms affected them in the last month. Patients are given

time before their consultation with the team to complete the form, some may even

complete the scale prior to their arrival for the evaluation. Any areas that are unclear to

the patient can be explained by a nurse or a staff member at the clinic familiar with the

scale. Once the patient has completed the RSI scale they take it with them into the team

consultation and the specialists review their symptoms before physical examination.

Clinical Rating Scale:

The presence and signs of reflux laryngitis are rated using the Reflux Finding Score

(Appendix C). This scale was developed by Belafsky et al. (2001b). The scale is utilised

during fiberoptic and stroboscopic laryngoscopic evaluation to gauge the presence or

absence of reflux based on clinical observation of typical signs. Observation is made

possible through the use of videostroboscopy. The team view structures within the larynx

and pharynx and rate them on the basis of presence or absence and severity of pathology.

This scale is independent of the patient and only the medical team conduct the rating. The

RFS is a consensus rating by three independent experts in voice, and was developed to

standardise laryngeal findings enabling clinicians to better diagnose and assess

therapeutic efficacy in patients with LPR (Belafsky et al., 2002).

36

Instrumentation:

The laryngoscopic evaluation for each patient was conducted using the Kay Elemetrics

9100B rhino-laryngeal stroboscope. This allowed the anatomical structure of the larynx

and pharynx to be observed while functioning, therefore, sanctioning the clinician to

make judgements related to any possible pathological structures seen and their impact on

functioning of the vocal folds. The stroboscopy is a crucial part of the measurement

procedure, as it enables the clinician to complete the RFS.

This form of evaluation allows detection of vibratory asymmetries, structural

abnormalities, small masses, submucousal scars, laryngeal neoplasims and other

conditions that are not visible under ordinary light (Sataloff, 2005a). The strobe

illuminates different points on consecutive vocal fold waves and the lighted portions are

fused together visually to create a cycle. The perception of the stroboscopy being in slow

motion is created by the strobe light and the frequency of vocal fold vibration

desynchronised by approximately 2 Hertz (Sataloff, 2005a). By combining the strobe

light with a camera these patients evaluations can be recorded and re-evaluated at a later

date or by other professionals. Sataloff (2005a) indicates that on initial examination his

practice routinely evaluates patients with both flexible and rigid endoscopes. The flexible

laryngoscope allows evaluation and documentation of the natural laryngeal motion

without obstruction from the tongue, as well as being useful in patients with a sensitive

gag reflex who, cannot tolerate rigid trans-oral endoscopic usage (Sataloff, 2005a).

Both flexible trans-nasal and rigid trans-oral endoscopic video-stroboscopy was

conducted. The trans-nasal flexible fiberscope allowed for observation of the upper vocal

tract and full evaluation of connected speech, laryngeal and supraglottal postures

(Morrison, Nichol & Rammage, 2001). The rigid fiberoptic scope was used orally to view

the same structures. Both trans-oral rigid fiberoptic endoscopes and trans-nasal flexible

fiberoptic endoscopes are used as they provide different information. The flexible trans-

nasal endoscope provides visualisation of the vocal folds with normal laryngeal posture

and provides a more accurate assessment of the posterior glottic opening, than during

rigid endoscope usage (Sataloff, 2005b). A rigid fiberoptic endoscope used trans-orally

provides a better more detailed picture which allows more intricate findings, as well as a

better 3 dimensional perspective of the vocal fold edge (Sataloff, 2005b).All patients

37

were examined using the same equipment, ensuring no differences or discrepancies in the

overall RFS rating between participants.

3.6 Procedure

A letter requesting permission to use the DGVS clinic as a site for research into

laryngopharyngeal reflux was submitted to the ENT’s who, work in the clinic. The

proposed area of research, rationale and implications for the study were presented in the

letter. A letter of approval from the site allowing access to their records was granted

(Appendix D).

The patients were previously evaluated at the clinic by an interdisciplinary team

composed of: two Otolaryngologists, a Speech Therapist and a Voice Coach. The

members of the team are experienced with voice patients and professional voice users.

The team members consult from the Donald Gordon Voice and Swallowing Clinic. At

each consultation with said team the participant completed a RSI form and the team

completed a RFS evaluation based on stroboscopy findings. Both the RFS and the RSI

scores were filed in the patient files at the clinic.

The researcher reviewed the files of patients seen at the clinic. The files of each patient

were dissected and the relevant information necessary for this study extracted. The RFS

and RSI totals for each patient’s first visit were examined and evaluated. An RSI score of

more than five (>5) indicates the possibility of Laryngopharyngeal Reflux, (Belafsky et

al., 2001b). Therefore, patients whom had a score of greater than 5 on the RSI scale were

included in the study. The totals for the RFS and the RSI and the score for each item on

the scales were captured on a spreadsheet thus constructing a database. Further,

information required to fulfil the sub-aims of the study including age, gender, profession

and smoking status were also included in the database. With respect to age, participants

were divided into subgroups of a 10 year age span producing 8 subgroups numbered 1 to

8. Patients were classed as either professional voice users (speech therapist, actor, singer,

performer, rabbi, teacher/lecturer) or non professional voice users.

38

Therefore each patient on the Microsoft Excel database had a set of RFS and RSI sub-

scores as well as total RFS and RSI scores. Gender, age, profession and smoking status

were also captured. Each patient’s information and data was allocated a file number

within a coding system instead of being correlated to a name. Therefore anonymity and

confidentiality of patient records was maintained.

From this data a summarised Microsoft Excel spreadsheet was constructed. The summary

was more concise and was formatted so that it could be used in the SAS program for

statistical analysis. The patients were coded as ‘s’ if they were smokers and ‘ns’ if non-

smokers. They were labelled as ‘m’ for male and ‘f’ for female, ‘y’ if they were

professional voice users and ‘n’ if they were non-professional voice users. Patients were

separated into age subgroups (1-8) as described above. Items on the RFS and the RSI

were abbreviated to an acronym in order to make the spreadsheet more succinct than the

original spreadsheet. An abbreviated example of an entry on the Excel spreadsheet is

provided below in figure 3.3.

# smoke Age gender Prof voice SO VO ER VCO DLO PCH GRAN OT Tot

2. ns 35 m n 0 0 4 3 3 3 2 2 17

Figure 3.3. An example of the coding system utilised in the excel spreadsheet.

Patient files were examined and analysed in sets of 20-30. The researcher made a list of

files that were being utilised and left this at the clinic, should it be required for follow up

visits by the patients or for reports.

The database was held by the researcher as it was a working document and was updated

as more patient files and scores were added to the database. Once the spreadsheet was

completed and all aspects had been coded, statistical analysis was conducted using the

SAS 9.2 analysis programme.

39

3.7 Data Analysis

The scores on the RFS and RSI scales are categorical variables that can only take on a

small set of values (Howell, 1999). Therefore, the scores are discrete points (0,1,2,3,4,5)

and are ordinal, providing no information regarding differences between points of the

scale (Howell, 1999). Ordinal data requires a nonparametric methods of analysis.

(Shiavetti & Metz, 2002). The total RSI and RFS scores could however, be treated as

continuous variables and analysed using parametric statistics. Specifically the Pearson

Correlation coefficient was used to establish the relationship between the RFS and the

RSI scores, the primary aim of the research.

The non-parametric Cohen’s Kappa co-efficient was used to correlate the sub-scores

within and between the RSI and RFS. However, this proved difficult as the RFS and the

RSI use varying methods of measuring the data i.e. some aspects on the RFS are scored

as absent or present using scores of 0 or 2 whereas others are rated in severity from 0 to 4.

The RSI uses a 5 point rating scale for all items. Where possible therefore, the correlation

was done automatically by the SAS 9.2 programme.

The second aim to ascertain which aspects of the RSI and the RFS are raised in the

patients was achieved using descriptive statistics.

The third aim was to describe trends related to the RFS and the RSI. This aim was also

achieved through descriptive statistics.

The fourth aim to verify if extraneous factors such as age, gender, professional voice use

and smoking impact on the totals of the RFS and RSI was achieved using t-tests for

gender, smoking status and professional voice use and an analysis of variance to assess

whether the age groups were significantly different from each other on the total RSI and

RFS scores.

The differences between groups on the sub-scores of the RSI and RFS were assessed

using the non parametric Chi–square statistic.

40

3.8 Reliability & Validity

Reliability:

The Rosenthal effect, specifically the interactional effect (i.e. the observer affects the

recording of the participant’s behaviour), observer effects, interpreter effects and

intentional effects (Shiavetti & Metz, 2002) are reduced in the current research through

the following methods. Observer and interpreter effects are reduced as the researcher and

the team conducting the assessments are not the same. The researcher is not present

during the rating of the RFS nor is the researcher present with the patient while

completing the RSI. As the team assessing the patient have no vested interest in the

outcome it is unlikely they will influence the RFS score either to indicate or not to

indicate reflux laryngitis. The fact that the researcher does not have any interaction with

the participants at any stage in the research further reduces observer and interpreter

effects. Intentional effects are accounted for by having more than one team member

assessing the participant’s clinical characteristics on the RFS. By doing so any

discrepancies between scores are discussed and a consensus regarding the score is

achieved and thus inter-rater reliability is achieved (Schiavetti & Metz, 2002).

Essentially three raters (2 otolaryngologists and 1 speech therapist) score each item on the

RFS together as a group, and a consensus between raters must be reached on each item of

the RFS before the result is recorded. Therefore the RFS’s reliability is realized.

Validity

The RFS is a nonreactive measure, it does not change what is being measured, and does

not impact on the internal validity of the study (Schiavetti & Metz, 2002). The Reflux

Severity Index is a reliable and validated measure of reflux signs. It is highly

reproducible and exhibits excellent construct-based and criterion-based validity (Belafsky

et al., 2002). Both the RFS and the RSI are well utilised scales in research (Joniau,

Bradshaw et al., 2007). The external validity and the ability to generalise the results of

this study is improved by increasing the sample size (Schiavetti & Metz, 2002).

41

Results

The results will be presented according, to the aims set out in the methodology.

Descriptive statistics will supplement the statistical results. The level of significance

chosen for the various statistical analyses was p=0,05 (5%), as the larger the sample the

smaller the level of significance selected (Dallal, 2003).

4. 1 Description of Results on the RFS and the RSI

The mean, mode, median, maximum and minimum for the RSI and RFS totals are

indicated in table 4.1 below.

Table 4.1 Summary statistics for the RFS and the RSI

Description RFS (max score = 26) RSI (max score = 45)

Mean 16,95 19,24 Mode 19 12

Median 18 19 Max 26 45 Min 1 6

Standard Deviation 4,73 8,08

The mean total rating of 16,95 on the RFS is indicative of LPR according, to Belafsky et

al. (2001b). The mean total rating of 19,23 on the RSI indicates LPR (Belafsky et al.,

2002).

42

Of the 105 participants, three were rated as presenting with signs reflecting the highest

score (26) on the RFS, while only one participant’s laryngeal signs were rated as a 1

which according, to Belafsky et al. (2001b) is not indicative of LPR. Only two

participants attained the lowest score considered to be indicative of LPR (=7) on the RFS.

The mode, which is the most frequent rating attained, on the RFS is 19, suggestive of

mild to moderate LPR. The median, the rating that is most central in the distribution is

18, also indicative of mild-to-moderate LPR. After the mode of 19, the next most

frequent rating was 17 in 9 participants. More detailed analysis of the number of

participants who attained total ratings in various categories on the RFS is shown in figure

4.1.

5

22

65

13

0

10

20

30

40

50

60

70

1 to 7 8 to 14 15 to 21 22 to 26

Total ratings on RFS 1 - 26

Nu

mb

er o

f p

art

icip

an

ts w

ho o

bta

ined

each

tota

l ra

tin

g

Figure 4.1 Number of participants who attained various total RFS scores.

The totals’ ratings on the RFS indicate that the majority of participants were rated

between 15 and 21 (65 participants) followed by a rating of between 8 and 14 (22

participants). The upper end of the RFS ratings was from 22 to 26 and only applied to 13

participants, while the lowest ratings between 1 and 7 applied to 5 participants. The exact

counts for each rating on the RFS are located below in table 4.2

43

Table 4.2 Exact ratings on the RFS and number of participants to whom each rating

applied.

RFS Rating Number of participants

1 1

6 1

7 3

8 2

9 1

10 4

11 1

12 3

13 4

14 7

15 6

RFS Rating

Number of participants

16 8

17 9

18 12

19 14

20 9

21 7

22 1

23 6

24 3

26 3

Further descriptions of the number of participants obtaining different totals on the RSI are

found in figure 4.2 below. The mode on the RSI is 12 while the median is 19. The

maximum rating on the RSI was 45 obtained by 1 person, while the minimum rating was

6.

44

15

26

18

21

16

5

2 2

0

5

10

15

20

25

30

6 t

o 1

0

11 t

o 1

5

16 t

o 2

0

21 t

o 2

5

26 t

o 3

0

31 t

o 3

5

36 t

o 4

0

41 t

o 4

5

Total RSI score range 6-45

Nu

mb

er o

f p

art

icip

an

ts w

ho o

bta

ined

each

tota

l ra

tin

g (

n=

105)

Figure 4.2 Number of participants who rated the total RSI as various severity scores.

Figure 4.2 depicts in which self-rating category most participants placed their symptoms

on the RSI. The higher ratings only applied to 4 participants whereas the range from 11 to

15 yielded the most participant ratings (26), closely followed by the 26 to 30 range (21),

then 16 to 20 (18), 26 to 30 (16), the lowest range between 6 and 10 applied to 15

participants and the 31 to 35 range applied to 5 participants.

45

Table 4.3 Number of participants to whom each rating on the RSI applied.

RSI Rating Number of participants

6 6

7 3

8 3

9 2

10 1

11 4

12 10

13 6

14 4

15 2

16 2

17 5

18 3

19 3

20 5

21 6

RSI Rating

Number of participants

22 5

23 2

24 3

25 5

26 4

27 6

28 1

29 2

30 3

33 1

34 2

35 2

37 1

38 1

43 1

45 1

Table 4.3 indicates the number of participants to whom each exact rating on the RSI

applied. The number of ratings over 30 steadily decline and only a handful of participants

had total RSI ratings over 35.

4. 2 Trends on RFS and RSI ratings

To establish which test items of the RSI and the RFS were most elevated in patients. The

cumulative total rating on each item for the 105 participants was divided by the total

possible maximum score yielding a percentage rating for each item.

The motivation for using a percentage as the best way to establish the most frequently

rated signs or symptoms in the sample is twofold. Firstly, the number of participants that

attained a rating on that item could not be used as each participant is rated on every item.

46

This means there would be 105 for each item (the total number of participants). Secondly,

the RFS is not a uniform scale of measurement unlike the RSI. The RFS has some items

that are quandary (have a point rating system) and others that are binary. The total rating

for each sign could not be used as a measure for highest rated sign as certain signs would

yield a higher total score based on the fundamental nature of the scoring utilised on the

RFS. By using percentage to establish which signs and symptoms were most elevated

provides a uniform measurement regardless of rating system.

The RFS sub-score trends were examined first. These are provided in table 4.4 below

which clearly indicates which laryngeal sign was rated the highest and was therefore the

most common laryngeal sign seen in the participants.

Table 4.4 Totals for each sign on the RFS and percentage of the total score

SO VO ER VCO DLO PCH Gran OT

Total cumulative recorded

score

46 190 378 299 317 342 67 148

Possible max total

score

210 420 420 420 420 420 210 210

Percent 21,9% 45,2% 90% 71,1% 75,4% 81, 4% 31,9% 70,4%

The most common laryngeal sign was erythemia (90%), followed by posterior commisure

hypertrophy (81,4%) and then by diffuse laryngeal oedema (75,4%). Erythemia is rated on

a three point scale. Of the 105 participants erythemia was rated as diffuse in 86 (81,90%)

participants and 17 (16,19%) participants were rated as having partial erythemia, while

only 2 (1,9%) participants were rated as having no erythemia present. The vast majority of

patients were rated as having diffuse erythemia.

Posterior commisure hypertrophy was rated as absent (0) in 3 (2,86%) participants, mild

(1) in 4 (3,81%) participants, moderate (2) in 14 (13,33%), severe (3) in 26 (24,76%) and

47

obstructing (4) in 58 (55,24%) participants. The majority of patients were thus rated as

having obstructing posterior commisure hypertrophy.

Diffuse laryngeal oedema (DLO) was the third most common laryngeal sign seen in the

participants’ laryngeal examination. DLO was rated as absent in one participant (0, 95%)

and mild (1) in one participant (0, 95%). A moderate rating (2) was assigned to 18 (17,

14%) participants while 60 (57, 14%) received a rating of 3, which is severe. Only 25 (23,

81%) participants were rated as having obstructing diffuse laryngeal oedema. Severe was

the most often assigned description to laryngeal oedema in the 105 participants.

Vocal Cord Oedema at any severity was seen in 103 participants and was the fourth

highest score (71, 1%) compared to other subtests on the RFS. This was closely followed

by “Other” (which indicates thick endolaryngeal mucous) 70,4%. Vocal cord oedema is

rated as absent to polypoid. The participants’ ratings revealed that 2 (1, 9%) were rated as

having no vocal cord oedema (VCO), 5 (4,7%) with mild VCO, 23 (21,9%) with moderate,

52 (49,52%) with severe and 23 (21,9%) participants with polypoid VCO. Therefore the

same number of participants who had moderate VCO had polypoid VCO. The majority of

participants had severe VCO.

Laryngeal mucus (“other”) was deemed to be absent in 31 (29,52%) participants and

present in 74 (70,48%) of them.

Ventricular oedema and granuloma were the next most common laryngeal signs seen and

lastly subglottic oedema was identified in only 21, 9% of participants. Of the total

participants 57 (54,29%) had partial ventricular oedema (VO), 18 (17,14%) had complete

VO and 28 had no VO. Granuloma was absent in the majority of participants (71) and

present in 33 of the sample. Subglottic oedema (SO) was present in 23 (21,9%) of the

records reviewed and absent in most, 82 (78,1%) of the sample.

The description of RSI sub-score trends was conducted in the same way as for the RFS and

are set out in table 4.5 located below.

48

Table 4.5 Totals for each sub-rating on the RSI and percentage of the total

HRS TC Muc SWD CLD BRD AC Lum HB

Total score for category

365 328 349 111 144 108 195 226 199

Possible total 525 525 525 525 525 525 525 525 525

Percent 69,5% 62,4% 66,4% 21,1% 27,4% 20,5% 37,1% 43,04% 37,9%

The total summary indicates that hoarseness was the most commonly self-rated symptom

by 69,5% of all participants, followed by mucous in 66,4% of participants and throat

clearing in 62,4%. The next most frequently rated symptom was globus sensation

(43,04%). Heartburn (37,9%) and annoying cough (37,1%) were very similarly rated in

terms of frequency of occurrence by participants. Coughing when lying down was rated as

less frequent by 27,4% of participants. Swallowing difficulties (21,1%) and breathing

difficulties (20,5%) were both rated as occurring less frequently than the other symptoms.

Below is a more in-depth look at each symptom and the number of participants for each

rating on that symptom.

49

Number of participants for each rating on Hoarseness

7

5

11

22

28

32

0

5

10

15

20

25

30

35

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f p

art

icip

an

ts (

n=

105)

Figure 4.3 Number of participants for each rating of hoarseness

It is clear that hoarseness is most commonly rated by 32 (30,48%) participants as severe

(5), 28 (26,67%) rated hoarseness as 4, 22 (20,95%) rated hoarseness as (3), 11 (10,48%)

rated this symptom as (2) and 5 (4,76%) rated hoarseness as (1) and 7 (6,67%) rated it as

no problem (0).

50

Number of partcipants for each rating of Mucus

5

10

13

21

30

26

0

5

10

15

20

25

30

35

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f p

art

icip

an

ts (

n=

105)

Figure 4.4 Number of participants for each rating of Mucus

Excess mucous was the second highest rated symptom on the RSI with 66, 4% of

participants 26 (24, 76%) rating it a (5), and 30 (28, 57%) rating this symptom as a (4).

The majority of participants felt this symptom was a (4) as opposed to severe (5). Of the

participants 21 (20%) rated mucous as (3), 13 (12, 38%) as (2), 10 (9, 52%) as (1) and 5

(4, 76%) as no problem. Therefore most patients feel that their mucous is not severe

enough to warrant being on the highest end of the rating scale.

51

Number of participants for each rating of Throat Clearing

9

6

15

29

25

21

0

5

10

15

20

25

30

35

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f p

art

icia

nts

(n

=105)

Figure 4.5 Number of participants for each rating of Throat Clearing

Throat clearing was rated as third most common symptom by participants with a score of

62,4%. Of the sample 21 (20%) rated throat clearing as severe (5), and 25 (23,8%) rated

throat clearing as relatively severe (4). The majority of participants 29 (27,62%) rated this

symptom as moderate (3), and 15 (14, 29%) assigned a rating of (2). The smallest group of

participants (6 i.e 5,71%) rated this as (1) while 9 (8,57%) of the sample rated throat

clearing as no problem.

52

Number of participants for each rating of Sensation of a Lump in

the throat

31

15

13

17

11

18

0

5

10

15

20

25

30

35

0 1 2 3 4 5

Ratings 0-5

Nu

mb

er o

f p

art

icip

an

ts

(n=

105)

Figure 4.6 Number of participants for each rating of Sensation of a Lump in the Throat

Lump in the throat (43,04 %,) was rated as having less impact on the participants. Lump in

the throat was rated by 31 of the participants (29,52%) as never occurring (0). Juxtaposing

this, was the next largest group of participants 18 (17,14%) who, rated lump in the throat

as severe (5). This symptom was rated by 17 (16,19%) participants as (3), 11 (10,48%)

rated this as (4) and 13 (12,38%) rated globus sensation as (2) while only 15 (14,29%)

rated this symptom as a (1).

53

Number of participants for each rating of Heartburn

37

11 11

24

11 11

0

5

10

15

20

25

30

35

40

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f p

art

icip

an

ts (

n=

105)

Figure 4.7 Number of participants for each rating of Heartburn

Heartburn was rated as having less impact on the total number of participants (37,9%).

Heartburn was rated as not a problem (0) by 37 (35,24%) participants, 11 (10,48%) of the

sample rated it as (1), (2), (4) and severe (5). Of the charts reviewed, 24 (22,86%)

indicated that heartburn was moderate (3).

54

Number of participants for each rating of Annoying Cough

34

18

15

18

7

13

0

5

10

15

20

25

30

35

40

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f p

art

icip

an

ts (

n =

105)

Figure 4.8 Number of Participants for each rating of Annoying cough

Annoying cough (37,1%) which according, to participants occurred less frequently than

other symptoms was rated by 34 (32,28%) as no problem (0), 18 (17,14%) rated it as (1),

15 (14,29%) of the sample experienced this symptom as a (2), 18 (17,14%) rated it as (3),

7 (6,67%) rated this symptom as relatively severe (4) and only 13 (12,38%) rated this as

severe (5).

55

Number of participants for each rating of Coughing when Lying

Down

49

17

12 13

68

0

10

20

30

40

50

60

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f P

art

icip

an

ts

(n=

105)

Figure 4.9 Number of participants for each rating of Cough when Lying down

Coughing while lying down was noted as occurring less often (27, 1%) than other

symptoms on the scale. The majority of participants 49 (46,67%) said coughing while

lying down never occurred, 17 (16,19%) rated this as mild (1), 12 (11,43%) as (2), 13

(12,38%) rated this as occurring sometimes (3), 6 (5,71%) rated it as a (4) and only 8

(7,62%) rated this as severe (5).

56

Number of participants for each rating of Swallowing difficulties

64

12

89

4

8

0

10

20

30

40

50

60

70

0 1 2 3 4 5

Rating 0-5

Nu

mb

er p

art

icip

an

ts (

n=

105)

Figure 4.10 Number of participants for each rating of Swallowing Difficulties

Swallowing difficulty occurred less frequently only 21,1% compared to other items on the

RSI. Swallowing difficulties were rated by most participants as never occurring (0).

Swallowing difficulty was rated as severe (5) by only 8 (7,62%) people, 4 (3,81%) rated

the symptom as relatively severe (4), 9 (8,57%) rated it as a (3), 8 (7,62%) rated this as a

(2), 12 (11,43%) rated this symptom as (1) and the majority 64 (60,95%) more than half,

rated this as never occurring (0).

57

Number of participants for each rating of Breathing Difficulties

64

8

1312

2

6

0

10

20

30

40

50

60

70

0 1 2 3 4 5

Rating 0-5

Nu

mb

er o

f p

art

icip

an

ts (

n=

105)

Figure 4.11 Number of participants for each rating of Breathing Difficulties

Breathing difficulties were rated according, to the charts reviewed as being the least

common symptom for LPR (20,5%). Breathing difficulty was rated as severe (5) in only 6

(5,71%) participants, 2 (1,90%) rated themselves as a (4), 12 (11,43%) rated this as a (3),

13 (12,38%) rated this a (2) and 8 (7,62%) rated this as a (1), while the overwhelming

majority 64 (60,95%) of 105 participants rated this symptom as never occurring (0).

4. 3 Correlation of the total RFS and RSI ratings, and inter-item and intra-scale

correlations:

A comparison of each of the total scores on the RSI and the RFS for each participant was

conducted. The comparison revealed that the participants (18) who, were rated as

demonstrating mild LPR on the RFS, had total RSI ratings of 22 and 11 respectively. The

three participants whose LPR signs warranted a total rating of 26 (severe LPR) on the

RFS, subjectively rated their symptoms on the RSI as 12, 19 and 23. Only one participant

rated her symptoms as 45 on the RSI, which is the highest total rating, yet her RFS rating

58

was only 10, which is indicative of mild LPR on the RFS. The minimum score on the RSI

(6) was only seen in six participants, and these participants were considered to have

moderately severe signs on the RFS (18, 6, 18, 12, 19, and 23)

Thus in general, there appears to be a lack of agreement between the subjective RSI rating

scale and the more objective RFS rating scale, and the subjective perception of symptoms

on the RSI are often more or less severe than the signs observed by the team of specialists.

Statistically this relationship was analysed using a Pearson correlation coefficient in order

to determine if a significant relationship exists between the two measures of LPR. The

results are set out in table 4.6.

Table 4.6 Correlation matrix for the total RFS and total RSI

RFS total RSI total

RFS 1.00000 -0.20125 p=0.0395

RSI -0.20125 p=0.0395 1.00000

A statistically significant weak negative correlation emerged from the analysis (r = - 0, 20

& p= 0,0395). Therefore there is an inverse relationship between the total RFS and RSI

ratings. This implies that while one rating increases the other decreases. The mean RFS

and RSI ratings (seen in table 4.1) may offer a degree of confirmation of this finding in

that the RSI total mean (19,24%) is higher than RFS total mean (16,95%).

Cohen’s Kappa Correlation co-efficient was used to statistically evaluate the intra-

correlation of items on the RFS and the RSI as well as the inter-item correlation across the

two scales. However, it was not always possible to use this method due to the differences

in scoring individual items on the RFS and the differences between the RFS and the RSI as

described in the methodology section. However, sub-scores that could be correlated were

done automatically in the SAS 9,2 programme

59

4. 4 RSI intra-item correlation

Table 4.7 Intra-item correlations on the RSI between hoarseness and other items

Variable Cohen’s

Kappa One-sided p Two-sided p

Significant

Correlation

Throat clearing 0, 2206 <0, 0001 <0, 0001 Yes

Mucous 0, 1367 0, 0004 0, 0008 Yes

Annoying cough 0, 0739 0, 0217 0, 0434 Yes

Globus sensation 0, 1561 <0, 0001 <0, 0001 Yes

Heartburn 0, 0396 0, 1505 0, 3006 No

Swallowing

difficulty 0, 0099 0, 3648 0, 7296 No

Coughing when

lying down -0, 0333 0, 1504 0, 3007 No

Breathing

difficulties 0, 0225 0, 2144 0, 4289 No

The intra- item–correlations on the RSI revealed the following relationships:

Hoarseness had a significant positive correlation with throat clearing, excess mucus,

annoying cough and lump in the throat. The relationship between hoarseness and heartburn

was not significant as was the case for swallowing difficulties, coughing when lying down

and breathing difficulties.

Values in table 4.8 show the intra-item correlation of throat clearing with other items on

the RSI.

60

Table 4.8 Intra-item correlations on the RSI between throat clearing and other items

Variable Cohen’s

Kappa One-sided p Two-sided p

Significant

Correlation

Hoarseness 0, 2206 <0, 0001 <0, 0001 Yes

Mucous 0, 2381 <0, 0001 <0, 0001 Yes

Annoying cough 0, 0633 0, 0503 - Yes (one-

sided)

Globus sensation 0, 1546 <0, 0001 <0, 0001 Yes

Heartburn 0, 0588 0, 0731 0, 3916 No

Swallowing

difficulty 0, 0038 0, 4505 0, 9010 No

Coughing when

lying down 0, 0189 0, 0088 0, 0176 Yes

Breathing

difficulties 0, 0086 0, 3916 0, 7832 No

Throat clearing was found to have a significant positive correlation with excess mucus,

coughing when lying down, feeling of lump in the throat and annoying cough (one-sided p

only). Non-significant correlations were found between throat clearing and swallowing

problems, heartburn and breathing difficulties.

It was established statistically that excess mucous had a significant positive relationship

with coughing while lying down (kappa = 0,0994; one-sided p = 0,0010 and two-sided p =

0,0020), annoying cough (kappa = 0,1208; one-sided p = 0,0005 and two-sided p = 0,

00010) and the sensation of a lump in the throat (kappa = 0,0948; one-sided p = 0,0076

and two-sided p = 0,0153) Non-significant correlation was noted between mucous and

breathing difficulties (kappa = -0,0092; one-sided p = 0,3693 and two-sided p = 0, 7386)

as well as swallowing difficulties (kappa = 0,0094; one-sided p = 0, 3685 and two-sided p

61

= 0,7371) and heartburn (kappa = 0,0573; one-sided p = 0,0637 and two-sided p = 0,1274).

The relationship between swallowing difficulties and coughing while lying down was a

strong positive correlation (kappa = 0,2321; one-sided p = < 0,0001 and two-sided p = < 0,

0001). This strong positive correlation was also seen between swallowing difficulties and

breathing difficulties (kappa = 0,3123; one-sided p = < 0,0001 and two-sided p = <

0,0001), annoying cough (kappa = 0,1185; one-sided p = 0,00047 and two-sided p = 0,

0093), feeling of lump in the throat (kappa = 0,1267; one-sided p = 0,0018 and two-sided p

= 0,0037) and heartburn (kappa =0,1303; one-sided p = 0,0022 and two-sided p = 0,0044).

A non-significant relationship was found between swallowing difficulties and mucous

(kappa = 0,0094; one-sided p = 0,3685 and two-sided p = 0,07371).

Coughing when lying down had strong positive correlation with breathing difficulties

(kappa = 0,3172; one-sided p = < 0,0001 and two-sided p = < 0,0001), annoying cough

(kappa = 0,3695; one-sided p = 0,0001 and two-side p = 0,0001), feeling of lump in the

throat (kappa = 0,1880; one-sided p = 0,0001 and two-sided p = 0,001) and heartburn

(kappa = 0,1527; one-sided p = 0,0006 and two-sided p = 0,0012).

Breathing difficulties had a strong positive correlation with annoying cough (kappa =

0,2319; one-sided p = < 0,0001 and two-sided p = < 0,0001), a feeling of lump in the

throat (kappa = 0,1019; one-sided p = 0,0096 and two-sided p = 0,0196) and heart burn

(kappa = 0,1391; one-sided p = 0,0014 and two-sided p = 0,0028). Furthermore a positive

significant correlation was found between annoying cough and feeling of lump in the

throat (kappa = 0,1849; one-sided p = < 0,0001 and two-sided p = < 0,0001), annoying

cough and heartburn (kappa = 0,1482; one-sided p = 0,0007 and two-sided p = 0,0014) and

heartburn and the sensation of a lump in the throat (kappa = 0,1213; one-sided p = 0,0040

and two-sided p = 0,0080).

Thus there were a number of significant intra-item correlations on the RSI, suggesting that

it is a reliable measure of LPR symptoms, as perceived by the patient, as this indicates

good content validity of this scale (Schiavetti & Metz, 2002).

An analysis of the co-occurrence of symptoms in terms of severity revealed interesting

results. The severity of many of the RSI items was closely associated. Severe hoarseness

(rated as 5) has a high association with severe throat clearing in 14 (13,33%) participants,

62

severe excess mucus (rated as 5) in 14 (13,33%) participants and severe sensation of a

lump in the throat (rated as 5) in 14 (13,33%) participants. Severe throat clearing has the

highest association to severe excess mucous in 15 (14,29%) of all participants. Severe

excess mucous is highly associated with severe throat clearing in 15 (14,29%) of

participants. Swallowing difficulties (rated as 3 which denotes that a symptom is

moderate) are associated with mild mucous (1) by 5 (4,76%) of participants and mild

swallowing difficulties (rated as 1); with mild coughing while lying down (rated as 1) by 5

(4,76%) of participants. Mild coughing when lying down (rated as 1) is associated with

infrequent throat clearing (2) in 7 (6,67%) of participants. Infrequent coughing while lying

down (2) was associated with infrequent annoying cough (2) by 7 (6,67%) of participants.

Breathing difficulties that are moderate (3) are associated with severe hoarseness (rated a

5) in 9 (8,57%) participants. Severe annoying cough (5) is associated with severe mucous

(5) by 10 (9,52%) of participants. Moderate heartburn (3) is associated with severe mucous

(5) by 11 (10,48%) of participants. A severe sensation of a lump in the throat (5) is

associated with severe hoarseness (5) by 14 (13,33%) of participants.

4. 5 Intra-item RFS correlation

Within the RFS intra-item correlation revealed the following:

Posterior commisure hypertrophy had a positive significant correlation with diffuse

laryngeal oedema (kappa = 0,2825; one –sided p = < 0,0001 and two-sided p = < 0,0001)

and ventricular obliteration (kappa = 0, 0622; one-sided p = 0,0391). A non-significant

relationship was seen between posterior commisure hypertrophy and vocal cord oedema

(kappa = 0,04008; one-sided p = 0,2215 and two-sided p= 0,4430).

A significant positive correlation was found between vocal cord oedema and diffuse

laryngeal oedema (kappa =0,1489; one-sided p = 0,0106 and two-sided p = 0,0212). Vocal

cord oedema and ventricular obliteration (kappa = - 0,0058; one-sided p = 0,4331 and two-

sided p = 0,8661) displays a non-significant negative correlation. Diffuse laryngeal

oedema and ventricular obliteration have a significant positive relationship (kappa =

0,0788; one-sided p = 0,0057 and two-sided p = 0,0115). “Other” and subglottic oedema

(kappa = 0,0554; one-sided p = 0,1772 and two-sided p = 0,3554) as well as granuloma

63

and erythemia (kappa = 0,0321; one-sided p = 0,1286 and two-sided p = 0,2571) displayed

a non significant correlation.

Table 4. 9 The intra-item association on the RFS: Severity levels at which the majority of

participants have co-occurrence between items.

VO (2)

SO (2)

ER (4)

VCO (3)

DLO (3)

PCH (4)

GRAN (2)

OT (2)

VO (2) 14 =

13,33%

51 = 48,57%

34 = 32,38%

37 = 35,24%

33 = 31,45%

19 = 18,10%

44 = 41,90%

SO (2) 14 = 13,33%

23 = 21,90%

10 = 9,52%

12 = 11,43%

17= 16,19%

8 = 7,62%

18 = 17,14%

ER (4) 51 = 48,47%

23 = 21,90%

47 = 44,76%

53 = 50,48%

55 = 52,38%

29 = 27,62%

67 = 63,81%

VCO (3)

34 = 32,38%

10 = 9,52%

47 = 44,76% 31 =

29,52% 34 =

32,38% 15 =

14,29% 41 =

39,05%

DLO (3)

37 = 35,24%

12 = 11,43%

53 = 50,48%

31 = 29,52%

32 = 30,48%

22 = 20,95%

40 = 43,81%

PCH (4)

33 = 31,43%

17 = 16,19%

55 = 52,38%

34 = 32,38%

32 = 30,48% 17 =

16,19% 48 =

45,71%

GRAN (2)

19 = 18,10%

8 = 7,62%

29 = 27,62%

15 = 14,29%

22 = 20,95%

17 = 16,19% 29 =

27,62%

OT (2) 44 = 41,90%

18 = 17,14%

67 = 63,81%

41 = 39,05%

46 = 43,81%

48 = 45,71%

29 = 27,62%

The above table highlights the severity levels at which the majority of participants have

co-occurrence between signs and symptoms. Ventricular obliteration (2) has the highest

co-occurrence with diffuse erythemia (ER), as 51 (48,47%) of all participants with this

sign, also had ER.

64

Diffuse erythemia was also found to have the highest association with subglottic oedema

in 23 (21,90%) participants, vocal cord oedema in 47 (44,76%) participants, severe diffuse

laryngeal oedema in 53 (50,48%) participants, obstructing posterior commisure

hypertrophy in 55 (52,37%) participants, granuloma in 29 (27,62%) participants and

“other” in 67 (63,81%) participants.

4. 6 RFS and RSI inter-item correlation

The between item analysis on the RSI and RFS was not possible using a Cohen’s’ Kappa

coefficient. However, the between item co-occurrence/association of the two scales was

evaluated. This was done by identifying the percentage and severity level at which the

majority of participants had rated themselves (RSI) and had been rated (RFS). Results of

this evaluation are found below in table 4.10

65

Table 4.10 Inter-item association of the RFS and the RSI: Severity levels at which the

majority of participants have an association between items.

HRS TC MUC SW CL BR AC HB LUM

VO (2) 19 = 18,10% (4)

22 = 20,95% (3)

17 = 16,19% (4)

7 = 6,67%

(1)

8 = 7,62%

(3)

6 = 5,71%

(2)

12 = 11,43% (3)

15 = 14,29% (3)

12 = 11,43% (3)

SO (2) 7 =

6,67% (4)

5 = 4,76%

(3)

7 = 6,67%

(4)

2 = 1,90%

(5)

5 = 4,76%

(3)

4 = 3,81%

(3)

5 = 4,76%

(3)

6 = 5,71%

(3)

4 = 3,81%

(5)

ER (4) 25 = 23,81% (4)

27 = 25,71% (3)

23 = 21,90% (4)

12 = 11,43% (1)

15 = 14,29% (1)

11 = 10,48% (2)

15 = 14,29% (1)

19 = 18,10% (3)

15 = 14,29% (3)

VCO (3)

16 = 15,24% (4)

18 = 17,14% (3)

14 = 13,33% (4)

9 = 8,57%

(1)

10 = 9,52%

(1)

7 = 6,67% (2+3)

10 = 9,52%

(3)

14 = 13,33% (3)

12 = 11,43% (3)

DLO (3)

18 = 17,14% (5)

20 = 19,05% (3)

17 = 16,19% (5)

10 = 9,52%

(1)

8 = 7,62%

(1)

9 = 8,57%

(2)

10 = 9,52%

(1)

11 = 10,48% (3)

11 = 10,48% (5)

PCH (4)

17 = 16,19% (5)

17 = 16,19% (3)

16 = 15,24% (4)

10 = 9,52%

(1)

13 = 12,38% (1)

7 = 6,67%

(2)

11 = 10,48

% (1+2)

10 = 9,52%

(3)

10 = 9,52% (3+5)

GRAN (2)

10 = 9,52%

(3)

13 = 12,38% (3)

9 = 8,57%

(4)

6 = 5,71%

(1)

6 = 5,71%

(1)

6 = 5,71%

(2)

9 = 8,57%

(3)

9 = 8,57%

(3)

7 = 6,67%

(1)

OT (2) 22= 20,95% (4)

22 = 20,95% (3)

19= 18,10% (3+4)

9 = 8,57% (1)

12 = 11,43

% (1+3)

9 = 8,57% (2)

12 = 11,43

% (1+3)

18 = 17,14% (3)

11 = 10,48

% (1+3)

66

This table indicates that laryngeal symptoms of the RFS clearly have a propensity to be

associated with certain RSI symptoms more than others. Partial ventricular obliteration has

the highest association with a moderate amount of throat clearing (3), in that 22 (20,95%)

participants had VO and TC (3). Subglottic oedema was associated the most with

hoarseness in 7 participants (6,67%) and excess mucous in 7 participants or (6,67%).

Diffuse erythemia had the highest association with throat clearing (3) with 27 (25, 71%)

participants, the majority, displaying both the symptom and the laryngeal sign. Severe

vocal cord oedema (5) was also highly associated with throat clearing (3) with 17, 14% or

18 participants out of the sample experiencing this configuration of signs and symptoms.

Diffuse laryngeal oedema (5) was associated most with throat clearing (3) by 20 (19, 05%)

of participants. Obstructing posterior commisure hypertrophy had the highest association

with both severe hoarseness (5) in 17 (16, 19%) participants and moderate throat clearing

(3) in 17 (16, 19%) participants. Granuloma was highly associated with moderate throat

clearing (3) by 13 (12, 38%) participants. “Other” was associated most with relatively

severe hoarseness (4) in 22 (20, 95%) participants and moderate hoarseness (3) in 22 (20,

95%) participants.

The symptoms of the RSI are easily seen to be associated with certain laryngeal signs of

the RFS. Hoarseness (4) has the highest association with diffuse erythemia by 25 (23,81%)

of participants. Throat clearing (3) has the highest association with diffuse erythemia as 27

(25,71%) participants presented with this combination. Mucus (4) and diffuse erythemia

had a distinct association in 23 (21,90%) of the participants. Swallowing difficulties (1)

was associated with diffuse erythemia by 12 (11,43%) of participants; infrequent coughing

while lying down (1) in 15 (14,29%) participants, breathing difficulties (2) in 11 (10,48%)

and infrequent annoying cough (1) in 15 (14,29%). A moderate amount of heartburn (3) in

19 (18,10%) as well as a moderate feeling of a lump in the throat (3) in 15 (14,29%)

displayed a higher association to diffuse erythemia.

Tables 4.9 and 4.10 expose the co-occurrence and associations between the items on the

RFS and the RSI and the inter-item associations of the RSI scale and RFS measures. These

associations allow insight into how symptoms and signs are related but do not and should

not lead to inferences on causation.

67

4. 7 Effects of gender, age, smoking and professional voice use on the RFS and RSI

totals

Table 4.11 Summary statistics for the total RFS and RSI by gender, smoking status and

professional voice use.

Gender Smoker Prof voice user

Male Female Yes No Yes No

N 43 62 21 84 50 55

Mean

RFS 18,40 16,05 18,43 16,63 17,68 16,36

Mean

RSI 17,26 20,60 19,33 19,20 19,62 18,87

Summary statistics for the total RFS (table 4. 11) indicate insubstantial differences

between the means for males and females, as well as for smokers and non-smokers and

also professional and non-professional voice users.

Summary statistics for the total RSI (table 4. 11) reveal that the mean for males is

approximately 17 and 20 for females. Smokers and non-smokers have very similar mean

ratings on the RSI, as do professional and non-professional voice users. However, non-

professional voice users have a lower total RSI rating than professional voice users.

68

Table 4.12 Mean ratings for each age group on the RFS and the RSI

N Mean Age Range Mean N

9 6,01 12-19 7,03 9

20 17,80 20-29 18,71 20

20 19,16 30-39 21,19 20

15 16,97 40-49 14,95 15

17 17,25 50-59 15,54 17

15 13,76 60-69 13,76 15

6 6,91 70-79 18,66 6

RFS

3 15,33 80-89 22 3

RSI

The mean totals for the RFS and the RSI for each age range are indicated in table 4.12

below. The highest mean for the RFS was seen in the 30-to-39 and the RSI the 80-to-89

(RFS= 19,16; RSI= 22) year old age range .This was followed by the 20-29 year olds

(RFS= 17,80) and 30-39 (RSI= 21,19) year olds. Then the 50-59 year olds (RFS= 17,25;

and 20-to-29 year olds RSI= 18,71. The lowest means for the RSI were seen in the lower

and upper end of the spectrum as was fewer participants for the age ranges. Interestingly

the 80-89 year olds had the highest RSI rating indicating they perceive their symptoms as

more severe.

ANOVA was used to assess the effects of age on the RFS and RSI totals. Participants

were divided into 8 age groups. There were no significant differences between age groups

on the total rating of the RFS (F = 1,48 and p = 0,1850) or the RSI (F = 0,45 and p =

0,867).

Using t-tests, the mean total ratings on the RSI and the RFS were compared to establish if

there are any significant differences between genders, and whether smoking status and

professional voice use affected these ratings significantly.

69

Smoking status did not have any significance on the total RFS assuming both equal (t= -

1,57; p= 0,1197) and unequal variance (t= -1,76; p= 0,0870) nor did it have an impact on

the total score of the RSI at the 5% level for equal (t= -0,06; p= 0, 9512) and unequal (t= -

0,08; p= 0,9403) tests of variance.

The gender comparison revealed interesting results. There were significant differences

between males and females on the RFS and RSI total ratings. Both tests for variance, equal

an unequal, were examined. Both the unequal tests of variance (t = -2,57, p = 0,0119) and

equal tests of variance (t = -2,61, p = 0,0105) yielded a significant difference on the RFS.

Gender showed a significant effect on the RSI total in the unequal test of variance only (t=

2,06 and p = 0,0414). Therefore gender does have an impact on the total RFS and RSI

ratings.

Professional voice use does not have any impact on the total of the RFS for both tests of

variance: equal (t = -1,43 and p = 0,1552) and unequal (t= -1,44 and p= 0,1520). The total

of the RSI is also not impacted by professional voice use for equal (t=-0,44; p= 0,6627)

and unequal (t= -0,44; p= 0,6609) tests of variance.

The Chi-squared statistic was used as a nonparametric measure to analyse the difference

between these variables on individual items of the RSI and the RFS. The only significant

differences were between age groups on throat clearing (Chi value = 64,469 and p =

0,0018) and coughing when lying down (chi value = 51,7848 and p = 0,0336); and

between males and females on subglottic oedema (chi value = 4,8315 and p = 0,0279) and

thick endolaryngeal mucous (chi value = 4,1728 and p = 0,0411).

4. 8 Trends related to variables of smoking, gender, age and professional voice use:

Despite the fact that there were very few statistically significant differences related to age,

gender, smoking status and professional voice on the individual items of the RSI and RFS,

specific trends were noted.

70

RFS trends related to smoking, gender, age and professional voice use:

The severity rating on each item of the RFS in each sample group is described. Subglottic

oedema (SO) was rated as present in smokers more frequently than it was in non-smokers

(28,57% vs. 24%). Participants in the age range of 70-79 (50%), 80-89 (33,3%) and 30-39

(26,32%) were more inclined to have SO. Males tended to be rated with SO more than

females (32,56% vs. 14,52%). Professional voice users were marginally more inclined to

be rated as having SO than non-professional voice users (24% vs. 20%).

Ventricular obliteration was rated as complete more often in smokers (28,57%) than in

non-smokers (14,29%). The age ranges from 70-79 (33,3%) and 80-89 (33,3%) as well as

50-59 (29,45%) were rated as having complete VO more commonly than the other age

ranges. Females and males had similar ratings for ventricular obliteration (16,13% vs. 18

6%). Professional voice users were rated as having VO in 18% of cases compared to 16%

of non-professional voice users. This is not a major difference between groups and

represents only a small proportion of both groups.

Erythemia (ER) was rated as diffuse in more smokers (90%) than non-smokers (79,9%).

Participants from 70-79 years (100%), 50-59 years (94, 12%) and 40-49 years (93,75%)

were rated as having erythemia more than the other age ranges. Males were rated with

erythemia in 90% of cases and females in only 75% of cases. ER was diffuse in 78,18% of

non-professional voice users and 86% of professional voice users. Therefore erythemia is

more prevalent in professional voice users within this sample.

Vocal cord oedema was rated as being severe in 61% of smokers and 46% of non smokers,

while 80-89 year olds’ (33,3%), 50-59 year olds (29,41%) and 20-29 years olds (28,57%)

were rated as having severe VCO more than the other age ranges. More men (60,46%)

were rated as having severe vocal cord oedema than women (41,94%). This symptom was

rated as polypoid in 28% of professional voice users and in 16, 36% of non-professional

voice users.

71

Diffuse laryngeal oedema was rated as obstructing in 38,10% of smokers and 20,24% of

non smokers. DLO was also seen in participants in the age ranges of 70-79 (66,6%), 60-69

(42,86%) and 50-59 (29,41%) more than the other ages. Obstructing DLO was rated

substantially more frequent in men (32,56%) than in women (17,74%), while 20% of

professional voice users and 27.27% of non-professional voice users were rated as having

obstructing oedema.

Post commisure hypertrophy was substantially more severe in smokers (71,3%) than non-

smokers (51,19%). Patients in ages 80-89 (66,67%), 70-79 (66,67%) ad 50-59 (64,71%)

had higher rating of PCH than the other ages. Of all the men 60% were rated as having

obstructing PCH compared to 51% of the female participants. Non-professional voice

users were rated with obstructing PCH in 50,91% of cases and 60% in professional voice

users.

Granuloma was rated as present more in non-smokers (34,52%) than smokers (19,05%).

The ages of 70-79 (50%), 40-49 (43,75%) and 50-to-59 (35,29%) were rated as having

granuloma more than other participant ages. This sign was rated as present in 41% of men

and only 24% of women. Granuloma was rated as present in 29,09% of non-professionals

and 34% of professional voice users.

Thick endolaryngeal mucous was rated as present almost equally in smokers (71%) and

non smokers (70%). The age ranges of 70-79 year olds (100%), 50-59 (76,46%) and 60-69

years olds (71,43%) were rated as having this symptom more often than the other age

ranges. Thick endolaryngeal mucus was rated as present in 81,40% of men and 62,90% of

females. Professional voice users were rated as having thick endolaryngeal mucus in 76%

of cases and non-professional voice users in 65,45% of cases.

Summary of RFS trends

Smokers were rated more frequently with the laryngeal signs of subglottic oedema,

erythemia, diffuse laryngeal oedema, posterior commisure hypertrophy and vocal cord

oedema, while non-smokers were more likely to present with granuloma. Thick

72

endolaryngeal mucus was rated as present in both smokers and non-smokers who,

participated in the study.

Male patients’ laryngeal signs were rated as more severe compared to females with respect

to subglottic oedema, erythemia, vocal cord oedema, diffuse laryngeal oedema, post

commisure hypertrophy, granuloma and thick endolaryngeal mucous. Males and females

had similar severity ratings for ventricular obliteration. Males on the whole appeared to

have more severe ratings than females on the RFS items.

Age trends related to the RFS reveal that the age group from 70-79 years were rated as

having the most severe laryngeal signs on six of the eight items on the RFS. These were

subglottic oedema, ventricular obliteration, erythemia, diffuse laryngeal oedema,

granuloma and thick endolaryngeal mucous. The age group from 80-89 was the group with

the next highest number of severe ratings. They were rated as most severe in one of the

eight laryngeal signs (vocal cord oedema). The 70-79 and 80-89 year old age groups

ranked the same on severity ratings for posterior commisure hypertrophy. On these eight

items, the 50-59 year old age group also had the most severe ratings on seven of the eight

items. Therefore age groups that seem to have a higher incidence of severe LPR ratings on

the RFS are: 70-79 year olds, 80-89 year olds and 50-59 year olds.

Professional voice users were rated as more severe than non-professional voice users on

selected items of the RFS, namely posterior commisure hypertrophy, thick endolaryngeal

mucus, granuloma, vocal cord oedema, erythemia, subglottic oedema. Professional and

non professional voice users had very similar ratings for ventricular oedema. Non

professional voice users were rated higher on diffuse laryngeal oedema.

RSI trends related to smoking, gender, age and professional voice use:

The highest rating on each RSI symptom was investigated and the percentage of

occurrence in the different populations noted.

73

Hoarseness was rated as severe (5) by 47,62% of smokers compared to 26,19% of non

smokers. Ages ranges of 40-49 (50%), 80-89 (33,3%) and 30-39 (31,58%) rated this

symptom as severe (5). Significantly more females than males rated themselves as a (5) for

this symptom (37,10% vs. 20,93%).

Throat clearing (TC) was rated as severe (5) by 21,43% of non-smokers and 14,29% of

smokers. Age ranges of 30-39 years (36,84%), 12-19 years (33,3%) and 80-89 years

(33,3%) rated the symptom of TC as (5). TC was rated as severe by more females

(25,81%) than males (11,63%). Professional voice users and non-professional voice users

displayed similar self ratings for severity of throat clearing (22% vs. 18,18%).

Similar numbers of non smokers (25%) rated excess mucous as severe (5) compared to

smokers (23%), however, more smokers (38%) rated mucous as a (4) compared to non

smokers (26%). Smokers may have different standards for excess mucous. Since smoking

is related to excess mucous production smokers may therefore not see this symptom as a

specific problem. The ages from 80-89 years (66,6%), 20-29 years (33,3%) and 30-39

years (31,58%) rated excess mucous as more frequent than the other age groups. Excess

mucus was rated as severe (5) by more females (29,03%) than males (18,60%). Excess

mucus was rated as severe by 18,18 % of non-professional voice users and 32% of

professional voice users.

Severe swallowing difficulties were rated very similarly by smokers (9,52%) and non-

smokers (7,14%). The ages of 80-89 years (33,3%), 70-79 years (33,3%) and 50-59 years

(11,76%) rated swallowing difficulties as severe (5). Gender comparison indicates that

males and females are fairly similar in their rating of this symptom as severe (9,30% and

6,45%). Swallowing difficulties were rated by 12,73% of non-professional voice users and

2% of professional voice users as severe (5).

Coughing while lying down was rated by most smokers (47,62%) and non-smokers (46%)

as never occurring (0). Participants from 80-89 years (33,3%), 40-49 years (18,75%) and

30-39 years (15,79%) rated this symptom as more severe (5) than other age groups.

Females (9,57%) rated this symptom as severe (5) whereas only (4,65%) of men rated it as

such (5). This is not a substantial difference but more females rated this symptom as

74

severe. Professional (6%) and non professional voice (9,09%) users were similar in their

rating of this symptom as severe.

Breathing difficulties or choking episodes were rated as severe (5) by more non-smokers

(7,14%) than smokers (0%). This may again be attributed to smokers having a higher

tolerance for coughing and choking due to the effects of smoking on the respiratory

system. Participants from 80-89 years (33,3%), 70-79 years (16,67%) and 60-69 years

(14,29%) rated this symptom as more severe (5) than other age groups. Breathing

difficulties were rated as severe (5) by (8,06%) of females and (2,33%) of males. A small

number (2%) of professional voice users rated this symptom as severe (5), while (9,09 %)

of non-professional voice users rated it as severe. Non-professionals therefore experienced

this symptom more frequently than professional voice users.

Annoying cough was rated by more non-smokers (20,24%) as severe (5) than by smokers

(4,76%). AC was rated by participants 80-89 (33 3%), 60-69 (21 43%) and 50-59

(17,65%) years of age as more severe than the other age ranges. Females (19,35%) rated

this symptom as a (5) substantially more than males (2, 33%). Annoying cough was rated

as severe by more professional voice users (18%) compared to non professional voice

users (7,27%).

Sensation of a lump in the throat was rated as severe by more non-smokers (17,86%) than

smokers (9,52%). Ages from 30-39 years (26,32%), 60-69 years (21,43%) and 40-49 years

(18,75%) rated this as severe (5) more than the other age ranges. Of females (22%) rated

this symptom as severe (5) whereas only (9,3%) of males did so. There was not much of a

difference between non-professional (16, 36%) and professional voice users’ (18,00%)

ratings of severe (5) for this symptom.

Heartburn is rated as severe by fewer smokers (14,29%) than non-smokers (25%).

Participants aged 80-89 (33,3%), 50-59 (23,53%) and 60-69 (21,43%) rated heartburn as

severe more than the other age groups. 16 (13%) of females and only (2.33%) of males

rated heartburn as severe. Heartburn was rated by (14,55%) of non-professional voice

users and only (6%) of professional voice users severe (5).

75

Summary of RSI trends

Smokers perceived their symptoms of excess mucus and hoarseness as more severe than

non-smokers. Non-smokers perceived their symptoms of lump in the throat, heartburn,

breathing difficulties, throat clearing and annoying cough as more severe than smokers do.

Smokers and non-smokers experience a similar severity of swallowing difficulties and

coughing while lying down.

Gender trends indicate that females rated symptoms of hoarseness, throat clearing, excess

mucus, coughing when lying down, annoying cough, breathing difficulties, a sensation of a

lump in the throat and heartburn as more severe than the male participants did. Females

and males seem to have similar severity ratings for swallowing difficulties. Females

seemed to perceive their symptoms as more severe than male participants did.

Age and RSI symptom related trends reveal that the age group from 40-49 years rated their

symptoms the most severe (5) in one out of the 9 items on the RSI. The age group 30-39

years rated their symptoms the most severe (5) in 2 out of the 9 RSI items and 80-89 year

old people rated their symptoms as the most severe (5) in 6 out of the nine RSI items. Ages

70-79 years, 60-69 years, 50-59 years and 20-29 year olds also featured as the top three

ages to have the highest ratings on RSI symptoms.

Professional voice users rated their symptoms of hoarseness, throat clearing, excess mucus

and annoying cough as more severe than non-professional voice users. Sensation of a lump

in the throat was rated similarly by professional and non-professional voice users, although

professionals rated this symptom as slightly more severe. Non-professional voice users

rated swallowing difficulties, breathing difficulties and heartburn as more severe.

Coughing while lying down was rated similarly by professional and non-professional

voice users, although non-professionals rated it as marginally more severe.

76

Discussion and Conclusion

The literature review conducted at the beginning of this study showed that the diagnosis

of LPR has been hotly debated. Consequently in some studies current methods of

diagnosis have been reviewed, re-examined and compared, while in others new

diagnostic methods have been proposed and their validity and reliability investigated. The

investigation undertaken here is representative of the former group and was aimed at

validating the specificity of diagnostic materials used for LPR.

The larynx is rendered more susceptible to acid and non acid reflux damage, than the

oesophagus, due to its histology and so the importance of a consistent and reliable

method for LPR diagnosis is crucial, especially when the long term medical repercussions

(e.g. cancer of the larynx) of LPR are taken into account.

The results from this study reveal a statistically significant but weak negative correlation

between the total RFS and RSI rating scales. As one variable changes it does in fact effect

change in the other. The negative correlation indicates that as the ratings on one scale

decreased, the ratings on the other increased. The findings also delineate trends related to

individual items on the RFS and RSI, as well as intra-item and between item correlations

and associations. The impact of other variables such as age, gender, smoking status and

professional voice use on the total RFS and RSI ratings also emerged.

Interpretation of Results

The significant weak negative correlation (r =- 0,20 & p= 0,0395) between the RFS and

RSI total ratings is confirmed by the means on the RFS and the RSI totals. As the RFS

ratings decrease the RSI ratings increase. This may be due to a number of factors and

impacting variables that will be discussed later.

Previous Studies

The results of this study are both confirmed and contradicted by similar studies such as

one by Mesallam, Stemple, Sobeih, and Elluru (2007). Messallam et al. (2007)

77

conducted a retrospective chart review of the RFS and RSI scores of 40 patients. They

found a significant strong positive correlation between the total RFS and RSI score (r =

0,86; p < 0001).

Vazquez, Fernandez and Gomez (2007) completed a study to evaluate the correlation

between the RFS and the RSI. They recruited a sample of 34 volunteers, all of whom had

no prior history of voice disorders. Each member of the sample was required to complete

a self-administered laryngeal symptom questionnaire (an amended RSI). Following this

all the participants underwent a comprehensive transnasal laryngoscopy in order to

document the reflux finding score. The results of their study concluded that there was a

statistically positive significant correlation between the RFS and the RSI. Furthermore

their study indicated that the correlation is greater when the total RFS is above 7.

Further, investigations into correlation of the total RFS and the total RSI was conducted

in a study of fifty–six laryngitis patients who were divided into two sub-groups, reflux

laryngitis and chronic laryngitis as well as 15 healthy subjects (Wang, Liu, Liu, Zeng,

Wu, Yang, Shen & Li, 2010). An oral and hypopharyngeal secretion was collected from

said patients and an assay was used to measure the pepsin concentration. This study

found that the total RFS and the total RSI were rated more severely (higher scores) in the

laryngitis patients who were part of the pepsin-positive group.

In contrast to this, a German study by Gugatschka, Schoekler, Kiesler & Friedrich (2008)

evaluated signs of the laryngeal tissue irritation through laryngoscopy (using the RFS)

and reflux associated complaints (using the RSI) in a cohort of 64 “healthy” male

volunteers. The researchers further wanted to describe if and to which grade subjective

symptoms correlate with clinical signs. Results of the study were that one third of the

patients presented with symptomology that was pathological on the RSI. Laryngoscopy

found laryngeal signs that could be attributed to LPR in only 6 of the participants. They

revealed further that symptoms and laryngoscopic findings did not correlate.

Musser et al. (2010) were also unable to find a statistically significant correlation between

the abnormal findings of the RFS (>7) and abnormal findings on the RSI (>13).

78

Current Study

The weak negative correlation in this study implies that as the RSI increases the RFS

decreases. This can be explained in a number of ways:

This suggests that laryngeal symptoms of LPR are experienced as more severe than

reflected by the actual examination of signs. This is not unusual as patients often

experience their symptoms of illness as more severe than they actually are, and given the

significance of voice for overall wellbeing, it is possible that this phenomenon is applies

to laryngeal pathology more than other pathologies. It is also possible that the laryngeal

signs of LPR develop more slowly than the symptoms do. This may be particularly true

of participants that have already undergone some degree of medical management from

other medical doctors. This information was not factored into the current study, and it is

unknown whether these participants had already had a trial medical management from

their general practioners.

Further issues such as professional voice users’ increased awareness of symptoms may

impact on their subjective ratings on the RSI, resulting in higher ratings than reflected on

the RFS. In addition, gender differences in ratings on the RSI may be due to

fundamentally different psychological ideologies of disease and the disease process.

These factors will be evaluated in more detail.

The way the RFS and the RSI were developed may also be a factor influencing the

correlation. The RFS and the RSI were created in tandem. The symptoms are a reflection

of the underlying laryngeal signs commonly seen in LPR. The 8 items on the RFS were

derived from a pool of the most common laryngeal findings of patients with LPR seen at

a voice clinic. That is, the creators (Belafasky et al., 2001b & Belafasky et al., 2002)

based the symptom ratings on the laryngeal findings and proposed that these findings

should result in specific symptoms. Therefore it is reasonable to assume that certain

symptoms would be more common with certain laryngeal signs.

Results from within and between-item correlations do indicate that there more within-

item correlations on the RSI than on the RFS.

79

For example, the perceived severity of a sensation of a lump in the throat was correlated

significantly with the perceived severity of all the other 8 items on the RSI. Globus

pharyngeus, or sensation of a lump in the throat, is according, to Mosca, Rossillo, &

Leone (2006) caused by cricopharyngeal muscle spasm. It is logical to conclude that if

one had a sensation of a lump in the throat one may experience increased symptoms of

throat clearing, annoying cough and coughing while lying down all in an effort to

dislodge the lump or ease the sensation. Perceived swallowing difficulties, breathing

difficulties and hoarseness may be increased by the ongoing feeling of a lump in the

throat. This is not to say that globus pharyngeus has a causative relationship with the

other symptoms, but an increased frequency in one will lead to an increased frequency in

the other.

Again, the symptom of annoying cough correlated significantly with: sensation of a lump

in the throat, heartburn, breathing difficulties, coughing while lying down, swallowing

difficulties, mucus, throat clearing and hoarseness. Cough in LPR is caused by airway

irritation due to the direct mechanism of contact with the refluxate and also gaseous

refluxate ( Mosca, Rosillo & Leone, 2006). It is reasonable to think that with an increase

in annoying cough a patient will have a perceived increase in the symptoms of breathing

difficulties, swallowing difficulties, throat clearing, hoarseness and coughing while lying

down. The more a patient coughs to try to clear mucus and the globus sensation the more

damage is done to the larynx in terms of oedema and erythemia therefore causing the

symptom of hoarseness to worsen.

Interestingly, hoarseness which had correlations at significant levels with only four items

on the scale: throat clearing; mucus; annoying cough and a sensation of a lump in the

throat, has very significant correlation with two symptoms – throat clearing and sensation

of a lump in the throat. Hoarseness is directly related to local irritation of the vocal cords

caused by refluxate returning to the hypopharynx (Mosca, Rosillo & Leone, 2006).

Reflux into the larynx stimulates increased mucus production as a defence mechanism by

the larynx. The excess thick endolaryngeal mucus provokes an increase in throat clearing,

which is considered vocal abuse (Sataloff, 2005b), because increased throat clearing

creates oedema and inflammations of the vocal cords as they are traumatically forced

together. Hoarseness is the consequence of this chain of events.

80

Coughing while lying down was correlated with all items except hoarseness. This is

because the increased mucus in the larynx creates a globus sensation and encourages

throat clearing and an annoying cough. This is especially true when reclining as the

mucus drains posteriorly. Excess mucus and oedema mean that the symptoms of throat

clearing, globus sensation, difficulty breathing, annoying cough and swallowing

difficulties become apparent. The diffuse laryngeal oedema and subglottic oedema may

shift posteriorly when the patient reclines thereby causing partial constriction of the

trachea and oesophagus leading to breathing and swallowing difficulties. Coughing when

lying down is not correlated with hoarseness as oedema in the vocal folds may have a

chance to dissipate slightly in the supine position allowing the folds to meet anteriorly. It

also seems unlikely that patients would be aware of the impact of hoarseness when lying

down as it is generally assumed that they would be retiring for the night.

Although mucus only correlates significantly with 4 items on the RSI (throat clearing,

annoying cough, hoarseness and globus sensation) it is one of the most commonly rated

symptoms and is one of the initiating factors in other symptom development. Mucus may

correlate less with the other symptoms (swallowing difficulties, breathing difficulties and

heartburn) as patients have already cleared the majority of mucus with throat clearing and

coughing before it has a chance to drain into the trachea and oesophagus in the upright

position. Therefore the mucus does not have the opportunity to be a causative agent of

these uncorrelated symptoms.

Swallowing problems correlated with 5 items on the RSI other than mucus, throat

clearing and hoarseness. Swallowing problems are generally directly related to diffuse

laryngeal oedema, erythemia and posterior commisure hypertrophy which will also cause

the manifestations of breathing difficulties, annoying cough, coughing while lying down,

and globus sensation.

Breathing difficulties are significant in LPR patients and are due to subglottal and

laryngeal oedema. The larynx is not able to perform one of its basic protection functions

of the airway when it is inflamed and swollen and consequently difficulties in breathing

may result.

81

Throat clearing correlated significantly with hoarseness, mucus, annoying cough and

globus sensation. This is important because of the relationship between these symptoms

and the complex interplay of one upon the other. Mucus goads throat clearing and

annoying cough which produce erythemia and oedema of the vocal cords and trigger

hoarseness.

A study mentioned previously by Wang et al. (2010) found similarly that the symptoms

of throat clearing, annoying coughing, and sensing a lump in the throat (P<0.006), were

more severe in the study group that had pepsin-positive assay than those who did not.

They therefore found that these symptoms were more prevalent in those patients who had

LPR. With the exception of the symptom of throat clearing, the current study had similar

findings. However, as discussed above these items are interlinked with each other due to

the patho-physiology of LPR. The symptoms are manifestations of the laryngeal signs.

The correlation between laryngeal signs on the RFS was not as abundant as the

correlations between symptoms on the RSI. Results of the Cohen’s kappa correlation

analysis indicate that only four laryngeal signs were significantly related. These were

diffuse laryngeal oedema, posterior commisure hypertrophy, vocal cord oedema and

ventricular obliteration. Posterior commisure hypertrophy appeared to correlate

significantly with only diffuse laryngeal oedema and ventricular obliteration. Diffuse

laryngeal oedema indicates that the entire larynx is swollen and therefore posterior

commisure hypertrophy is expected to correlate with this sign as posterior commisure

hypertrophy is caused by oedema. This is also the case with ventricular obliteration as

swelling in the larynx means that the ventricles are not easily distinguishable. Diffuse

laryngeal oedema correlated with only 2 other laryngeal signs – vocal cord oedema and

ventricular obliteration. Vocal cord oedema had significant correlations with ventricular

obliteration only, as swelling in the vocal folds leads to the ventricles being difficult to

differentiate. Ventricular obliteration appeared to correlate with the most number of signs

on the RFS, namely posterior commisure hypertrophy, vocal cord oedema and diffuse

laryngeal oedema. This is because posterior commisure hypertrophy (swelling posterior

to the commisure), vocal cord oedema (swelling of the vocal cords) and diffuse laryngeal

oedema (generalised swelling in the larynx) all obscure the visualisation of the ventricles

leading to their obliteration. The positive correlations seen in these laryngeal signs

82

essentially mean we can use one sign to predict the severity of another. If a patient is

rated highly on ventricular obliteration the rater can expect to see a more severe rating on

posterior commisure hypertrophy, vocal cord oedema and diffuse laryngeal oedema. As

with the symptoms of LPR the signs are highly interlinked.

The individual items on the RFS and the RSI could not be correlated statistically due to

reasons already discussed. The most common symptoms and signs in the cohort were

examined.

The total percentage for each symptom was calculated and the most common symptom

was hoarseness (69,5%) followed by excess mucus (66,4%) and then throat clearing

(62,4%). It is hardly surprising then that these are among the most commonly co-

occurring symptoms. These co-occurring items on the RSI were evaluated by examining

which sets of symptoms at all frequencies were self-rated by the highest proportion of

participants.

Remacle and Lawson (2006) suggest that the symptoms most related to LPR were throat

clearing, annoying cough, globus sensation and heartburn. In this study globus sensation

and heartburn were only the fourth and fifth most common symptoms. In a study by

Pribuisiene et al. (2002) a combination of three symptoms, hoarseness, throat itching, and

globus pharyngeus were found to separate LPR patients from healthy individuals, while

sore throat (40%), hoarseness (30%), and cough (20%) were the most common laryngeal

symptoms reported in a study by Qadeer et al. (2005). Another study found the most

important laryngopharyngeal symptoms were hoarseness in approximately 90% of the

sample, and globus sensation as well as chronic cough or difficulty swallowing (Rouev,

Chakarski, Doskov, Dimov, & Staykova, 2004). Thus, although there is some variation in

the symptoms listed as the most common in these studies, the current study has

hoarseness and annoying cough in congruence with these studies.

Hoarseness appears to be consistently seen as the most common symptom of LPR across

studies. Hoarseness may therefore be viewed as a consistent warning sign that clinicians

may use to alert them to the possible presence of LPR. However, hoarseness is a general

symptom that is the hallmark of many vocal pathologies e.g. PVFM and muscle tension

dysphonia. This idea was explored by Cohen & Garrett (2008) who found that patients

83

with persistent hoarseness may be over referred for trial PPI therapy (a trial PPI therapy

intervention is one method of diagnosis utilised in LPR), and that this may impact on

their treatment outcome. In other words hoarseness may be used as a means of

compartmentalising patients into LPR and non-LPR patients. By doing so patients may

fail to respond to PPI therapy as the patient does not have LPR. Therefore hoarseness

may be a red herring, a double edged sword. It may be that it is a strong indicator of LPR

whereas at the same time it may indicate other vocal pathologies. Thus, hoarseness

should be taken as only one possible symptom and should not be used to diagnose LPR

without further investigation.

Of all of the laryngeal signs, erythemia (90%) was rated the most severe in the total

sample of patients. This was followed by posterior commisure hypertrophy (81,4%) and

then vocal cord oedema (71,1%). This is not in agreement with Koufman, Amin &

Panetti (2000) who suggest that oedema is the most common sign used to diagnose LPR.

However, Remacle and Lawson (2006) indicated that physical examination mostly

revealed arytenoid erythemia and oedema, vocal cord erythemia and oedema, and

posterior commisure hypertrophy to be most related to LPR. Their findings are equivalent

to the findings from the current study. Another study reported that the most common

laryngeal signs in LPR were medial arytenoid wall erythemia/oedema (60%),

interarytenoid erythema (50%), and arytenoid complex erythemia/oedema (50%) (Qadeer

et al., 2005). These essentially mean erythemia and oedema are the most common

laryngeal findings in their study. This is the case in the current research project as well.

This is because reflux promotes inflammation and swelling of the larynx, which extends

to the vocal cords and the posterior commisure.

Commonly co-occurring laryngeal signs were examined by reviewing the severity of the

total set of signs. The set with the highest proportion of participants presenting with those

signs was considered to be the most commonly co-occurring set of signs. The severity of

each of these signs from the set was noted. The commonly co-occurring signs on the RFS

were not unexpected based on the highest/most rated laryngeal signs described in the

studies above. All laryngeal signs co-occurred with erythemia. This is the first sign of

tissue damage to develop from exposure to reflux and other signs develop from this one.

Therefore all the signs on the RFS will co-occur with erythemia.

84

Between-item correlation of the RFS and the RSI was not possible to due statistical

constraints. The scales are not equal types of measurement. The RSI is a 5 point scale

whereas the RFS has aspects that are binary and other aspects that are quandary (rated

from 1-4).

Musser, Kelchner, Neils-Strunjas, & Montrose, (2010) propose that the RFS where ‘0’

always represents ‘absent’ but all other scores represent something different between

scores makes rating difficult and not homogenous across raters. The RSI although having

a uniform rating of all items was criticised by Ali (2008) for not including throat pain,

which was found in up to 40% of LPR patients and for the inclusion of heartburn which is

a commonly reported symptom in esophageal reflux, not in LPR, and is known to

respond well to proton-pump inhibitors (PPIs). The researcher also questions the motives

for evaluating “cough” in two items (Ali, 2008). However, one rationale for the inclusion

of heartburn is perhaps to aid in differential diagnosis. Those who do have heartburn may

therefore be better able to be excluded from having LPR despite the RSI score.

A retrospective chart review of 40 randomly selected patients from a voice clinic in the

United States comprised the sample of a similar study. The charts were reviewed and

potential patients with signs of LPR included. All participants completed and RSI and the

RFS was completed post videostroboscopy. Between-item correlation of the RSI and the

RFS were calculated statistically and they report highly significant correlations between

signs and symptoms. Hoarseness correlated highly with vocal fold oedema (p <.01)

diffuse laryngeal oedema (p <.01) thick endolaryngeal mucus (p <.01) and erythemia (p

<.05). Throat clearing correlated with thick endolaryngeal mucus (p<.01) and vocal fold

oedema (p<.01). Globus sensation correlated with erythemia (p <.001) and posterior

commisure hypertrophy (p <.05) (Mesallam, Stemple, Sobeih & Elluru, 2007). The

statistical tests utilised in obtaining these results are unknown. However, it may be that

the researchers grouped or paired the data on the scales. The effects of this on the

reliability and validity of the scales is unclear, and it may detract from the significance of

the score when paired with another score.

A recent literature review of 383 articles related to LPR over the last thirty years was

undertaken by Kotby, Hassan, El-Makhzangy, Farahat, Shadi, and Milad (2010). This

review revealed that in 140 articles, 18 had used the incorrect statistical analysis

85

procedure. It is unclear if the study mentioned above by Mesallam et al. (2007) was

included in this review but it is likely, based on the date of publication.

Due to the differences in measurement of the RFS and the RSI inter-item correlation was

not possible statistically. Therefore links were made through examining the most

commonly co-occurring symptoms and signs on the two scales in a similar manner

described for inter-item co-occurrence. Co-occurrence between the scores indicates that

all the items on the RSI (symptoms) are seen the most often with the sign of erythemia.

Erythemia is inflammation and in this region it is the precursor to many of the signs that

follow, for example oedema. Therefore the symptoms of the RSI are seen with erythemia

as it is usually diffuse by the time patients’ experience symptoms related to laryngeal

oedema.

The laryngeal signs on the RFS were highly co-occurring with three core symptoms:

throat clearing, hoarseness, and excess mucus. This clearly defines the link between

laryngeal signs and symptoms. Reflux increases the amounts of mucus produced which

increases throat clearing which causes laryngeal irritation and oedema and ultimately

hoarseness.

Signs of ventricular obliteration, erythemia, vocal cord oedema and diffuse laryngeal

oedema were seen most often with an increased frequency of throat clearing. These

laryngeal signs and symptoms are interlinked. Erythemia and oedema are a result of

throat clearing as the patient tries to clear the excess mucus caused by refluxate.

Posterior commisure hypertrophy and thick endolaryngeal mucus occurred most

commonly with the symptoms of hoarseness and throat clearing. This is not surprising as

mucus and oedema, including posterior commisure hypertrophy, is the cause of throat

clearing and ultimately hoarseness. Participants who suffer from subglottic oedema were

found to have the symptoms of hoarseness and mucus as the highest symptoms compared

to others on the RSI. Those participants who had granuloma were found to have

hoarseness as the highest co-occurring symptom. This is because granuloma are growths

on the vocal cords and as a result this, incites hoarseness as the vocal folds are not able to

adduct completely. Therefore we can deduce that patients who present with laryngeal

signs of ventricular obliteration, erythemia, vocal cord oedema and diffuse laryngeal

86

oedema will in all likelihood present with an increased throat clearing rating. Those with

laryngeal signs of posterior commisure hypertrophy and thick endolaryngeal mucus may

have increased symptoms of hoarseness and throat clearing. The thick endolaryngeal

mucus goads throat clearing which is traumatic to the vocal cords and hoarseness is the

symptom that is experienced. Posterior commisure hypertrophy results from the reflux,

eryhthemia, and consequent oedema. Patients with raised ratings of granuloma may have

higher scores for symptoms of hoarseness as discussed above.

Clinically these connections may provide clinicians with some insight into what patients

may present with. Despite not being statistically relevant, we may have a better idea of

which symptoms and signs are more likely to be seen in combination.

In Table 4.10 the frequency with which each symptom is most commonly seen with

laryngeal signs and the severity level of that sign and symptom is highlighted. A direct

link can be drawn from the laryngeal sign to the symptom and the progression of LPR

and the frequency with which most patients experience particular combinations of signs

and symptoms. The same is true of table 4.4 and table 4.5 located in the results section. In

fact the most commonly seen signs and symptoms may be initial indicators of the

presence of LPR.

Trends

Other factors such as gender, age, smoking status and professional voice use were

examined to establish if they have any impact on the RFS and the RSI total scores as well

as on individual items. Very few studies that take into account, or at least discuss, the

impact of these variables on the RSI and the RFS, could be found.

Gender differences on the RSI indicate that females have higher self-ratings on the RSI

than males, since the mean total rating for females on the RS was 20,59% and for males it

was 17,25%. Females experienced all of the symptoms on the RSI as more severe than

males did. The symptom of swallowing difficulty was the only instance where males and

females had a very similar rating, not considered to be relevant.

87

These gender differences may indicate increased sensitivity and awareness of symptoms

in females or alternatively that males are reluctant to admit the actual severity of their

symptoms. Banks (2001) outlined in his article that men attend doctors less often than

females. Further men may view attending the general practitioner as a sign of weakness

and consequently look down on those who innconvenience the doctor with “minor”

symptoms (Banks, 2001). Males may in fact view their symptoms as minor and therefore

not weight them as heavily as they perhaps should. Females on the other hand may not be

hypersensitive to the symptoms, but may actually be reporting the severity of symptoms

more accurately than males do.

Smoking status in RSI symptoms revealed that smokers perceive the symptoms of excess

mucus and hoarseness as more severe than non-smokers. However, non-smokers

perceived their symptoms of breathing difficulties, throat clearing, annoying cough,

heartburn and sensation on a lump in the throat as more severe than non-smokers. Both

non-smokers and smokers had similar perceptions of the frequency of coughing while

lying down and swallowing difficulties. Non-smokers therefore perceive more of their

symptoms as worse than smokers. This may be due to the fact that smokers already have

many of the symptoms associated with LPR from smoking. Smoking is an irritant of both

respitory and laryngeal structures and has been linked to thinning of the mucosal lining of

these areas (Awan & Morrow, 2005). Due to this decrease in mucosal lining, some

smokers may present with an abusive cough as a result of increased sensitivity of the

structures. The study by Awan and Morrow (2005) elucidated further that smoking is

known to have effects on voice characteristics, such as fundamental frequency changes

and limitations in pitch range and is often connected to chronic laryngitis. Therefore some

of the symptoms listed on the RSI may in fact already be suffered by smokers and hence

they rate them as less severe than non-smokers who do not suffer from these symptoms

on a regular basis.

Age related RSI symptoms indicate that older participants in the age range 80-89 years

rated their symptoms the most severe on 6 of the 9 items. Participants in the age range

from 30-39 years rated their symptoms as the most severe on 2 of the 9 items and 40-49

years olds only rated one symptom as the most severe. However, participants in their

70’s, 60’s, 50’s and 20’s also featured in the top 3 age groups rating various symptoms as

88

severe. Clearly, participants in their 80th decade perceived their symptoms as more severe

than any other age group. This may be due to a number of reasons. Firstly, there were

only 3 participants in their 80’s. Therefore if any of the sample have a high score on an

item it skews the data upwards (or downwards if it is a low score) as one participant is

equivalent to 33,33% of the group. For this reason the researcher circumvented this small

sample dilemma by noting the top three age groups for each symptom, to provide a

clearer understanding of which age groups scored highest on each symptom. Secondly,

the participants in the 80-89 year old age group may have perceived their symptoms as

more severe due to other illnesses that have similar symptoms. It is well documented that

the symptoms used to diagnose LPR are not exclusive (Ford, 2005; Reimer & Bytzer,

2008). Elderly patients may also be on medication for concomitant disorders that have

similar symptoms or may make worse the symptoms the participant may already be

experiencing as a side effect. There are also a number of voice changes such as

hoarseness and changes in fundamental frequency that are related to age (Honjo, Isshiki,

1980). Patients should be counselled on these changes to ensure that the RSI is measuring

LPR and not age related symptoms.

Professional voice users rated their symptoms of hoarseness, throat clearing, mucus and

annoying cough as worse than non-professional voice users. Professional voice users

rated globus sensation as marginally more severe than non-professional voice users. Non-

professional voice users rated swallowing difficulties, breathing difficulties and heartburn

as more severe than professional voice users. While they had similar ratings for coughing

while lying down the non-professional voice users rated this as marginally more severe.

Overall the professional voice users rated more symptoms as severe than the non-

professionals did. In a study of 20 volunteer singing teachers, only 2 of the 20

complained of heartburn (linked to GERD). However, after completion of a questionnaire

it was found that 13 of the 20 had symptoms consistent with LPR (Heman-Ackah, Dean

& Sataloff, 2002), indicating that the singing teachers themselves may not be aware what

their symptoms suggest.

Another published study found that self-reported voice problems were twice as likely in

singing teachers as in controls (64% versus 33%) (Miller & Verdolini, 1995). When

singers and non-singers were asked to report on the history of any previously diagnosed

89

voice problems and if they had experienced any vocal disability within the last year, it

was found that singers were 1.7 times more likely to have had a history of vocal disability

in the preceding year (69% singers versus 41% non-singers) and singers were twice as

likely to have had a previously diagnosed voice problem when compared to the non-

singers(44% versus 21%,) (Phyland, Oates & Greenland, 1999).

Singers are more aware of vocal fluctuations and changes (Heman-Ackah, Dean, Sataloff

& 2002). They are also more aware of what changes in their voices may do to their

careers. For this reason, some professional voice users may rate their symptoms more

severely as a result of anxiety related to the voice changes that may impact their careers.

The impact of gender on the signs of LPR on the RFS indicate that males have higher

ratings for subglottic oedema, erythemia, vocal cord oedema, diffuse laryngeal oedema,

posterior commisure hypertrophy, granuloma and thick endolaryngeal mucus. Females

were not rated higher than males on any of the RFS items. Females and males had very

similar scores for one item on the scale, ventricular obliteration. Males are therefore rated

as having more severe signs of LPR than females do. This is confirmed when comparing

the total means on the RFS for males (18,39) and females (16,01). It is possible that

males only seek treatment later than females do, resulting in more severe laryngeal signs

later in the disease progression. An article on men and illness indicated that men with

health problems are less likely than women to have had recent contact with a doctor

regardless of income or ethnicity (Banks, 2001). This reluctance to seek medical attention

means that by the time they do seek attention the disease may have progressed

significantly. Late presentation can have serious consequences (Banks, 2001). Therefore

the males in this study in all likelihood fit the profile of males in general and probably

waited until the LPR symptoms had progressed further than the females did before

consulting a medical practioner. This would explain why males have higher RFS ratings

on all signs of LPR.

Sataloff (2005b) indicates that adult females tend to be more prone to vocal nodules and

Reinke’s Oedema. Reinke’s oedema is a mucoid gelatinous fluid in the superficial layer

of the lamina propria creating a polypoid appearance of the vocal fold (Sataloff, 2005a).

This increase in Reinke’s oedema in said population may be due to hormonal fluctuations

for example in peri-menopausal women (Sataloff, 2005b). Therefore, there is a chance

90

that hormonal differences and changes may impact other physiological sites to create

some of the signs assessed on the RFS. This is an area that should be studied further to

examine the effects of hormones on the laryngeal structures. This is also relevant in

considering the interaction between age and gender.

Smoking status is an important consideration when evaluating signs of laryngopharyngeal

reflux. After all, the oropharynx and laryngopharynx are the direct tissue sites for contact

with smoke. Results indicate that smokers were rated more severely on stroboscopic

examination than non-smokers on the following items: subglottic oedema, erythemia,

diffuse laryngeal oedema, posterior commisure hypertrophy and vocal cord oedema.

Interestingly non-smokers were only rated as more severe than smokers on granuloma.

Thick endolaryngeal mucus was rated almost equally severe in smokers and non-smokers.

The mean rating on the total RFS for smokers is 18,42 and for non-smokers is 16,63. This

difference indicates as the descriptive findings did, that smokers were rated with more

severe signs on the RFS than non-smokers.

A study that compared the videostoboscopic findings of smokers versus non-smokers

revealed that smokers have more observable signs of laryngeal irritation than non-

smokers (Awan & Morrow, 2005). Two key laryngeal characteristics related to smoking

have been identified as erythemia and oedema. The study stated that habitual smoking is

likely to promote thickening and possible oedema of the vocal folds. Further they

explained that erythemia may be restricted to the vocal folds or may be diffuse in the

surrounding laryngeal tissue. Awan and Morrow (2005) indicate that laryngeal exposure

to noxious agents such as cigarette smoke has been associated with chronic laryngitis and

erythemia. Fritzell and Hertegard (1986) as cited in Awan and Morrow (2005) reported

that 98% (123 out of 126) of patients who were found to have chronic vocal fold oedema

were smokers. The study found that even only relatively brief smoking habits can be

responsible to some degree of the laryngeal signs mentioned above.

The signs that were indicated to be directly impacted by smoking in the article by Awan

and Morrow (2005) were rated as more severe in this study for the smokers in the sample.

Furthermore many of these signs are also used as diagnostic criteria on the RFS for LPR.

It must be taken into account that smokers may therefore have another variable impacting

on the severity of their laryngeal scores and this should be taken into account when

91

evaluating patients who smoke for LPR on the RFS. This information on smoking

ultimately confirms that other airway irritation sources could be associated with

larynx/pharynx signs often attributed to reflux (Hicks et al., 2002).

Professional voice users had more severe ratings of posterior commisure hypertrophy,

thick endolaryngeal mucus, granuloma, vocal cord oedema, erythemia and subglottic

oedema. Non professional voice users had more severe ratings for diffuse laryngeal

oedema. Non professional and professional voice users had similar ratings for ventricular

obliteration. Overall professional voice users are rated as more severe in the majority of

items on the RFS than non professional voice users. The difference between the RFS

total means for professional (17,68) and non professional (16,36) voice users is small but

is skewed towards professional voice users.

A study into the laryngoscopic findings of singing teachers suggested that singers are

more likely to experience and be aware of subtle voice changes and suffer disability from

them. They may, as a result of this hyperawareness of their voice and how it fluctuates,

be more likely to seek treatment earlier than the non-professional voice user (Heman-

Ackah, Dean & Sataloff, 2002). Of the 20 singing teachers who volunteered for the study

7 complained of vocal issues while 13 were assigned to the ‘normal’ group based on no

vocal complaints. Despite the lack of symptoms, all of the volunteers had signs of reflux

laryngitis on examination. The signs included arytenoid oedema, arytenoid erythemia,

postcricoid oedema, interarytenoid pachyderma laryngis, and/or posterior cobblestoning.

Professional singers are widely known to have a higher incidence of vocal fold lesions

including nodules cysts and varices as well as an increased incidence of asymmetries in

vocal fold mobility (Heman-Ackah, Dean & Sataloff, 2002). In the study just mentioned

signs of LPR were found in all of the patients. According, to the authors these results

support the commonly held notion that laryngopharyngeal reflux may play an important

role in the development of laryngeal pathology in singers (Heman-Ackah, Dean &

Sataloff, 2002).

However, it is difficult to differentiate between the cause and the symptom. Are

professional voice users at increased risk for LPR or does LPR cause many of the signs

and symptoms seen in a population that may be susceptible to voice changes? Sataloff

(2005b) points out “the impact of laryngopharyngeal reflux disease in laryngology… is

92

most likely presently over-recognised and, therefore, over diagnosed and overtreated” p.

33. This is especially true in professional voice users.

Other research has demonstrated that a degree of LPR clinical signs have been found in a

large proportion of ‘normal’ healthy participants (e.g. Hicks et al., 2002). Their findings

on 105 normal volunteers whose inclusion in the study meant they did not have any signs

or symptoms related to reflux and no history of voice complaints, were revealing

regarding prevalence of LPR signs. Furthermore, their smoking status and history was

well documented with in-depth questions related to this. Their findings were that 86% of

the sample presented with at least one hypopharyngeal symptom (Hicks et al., 2002). The

fact that some normal people present with signs of LPR may lead one to conclude that

even ‘healthy’ voice users may have some symptoms of LPR.

The effect of age on the RFS indicated that the group from 70-79 years of age is one of

the top three age groups on ratings of subglottic oedema, ventricular obliteration,

eryhtemia, diffuse laryngeal oedema, granuloma and thick endolarygeal mucus. Those

from 50-59 years of age (50’s) were also one of the top three age groups. This group was

prevalent in six of the eight laryngeal signs, these were: erythemia, vocal cord oedema,

diffuse laryngeal oedema, posterior commisure hypertrophy, granuloma, thick

endolaryngeal mucus and ventricular obliteration.

Those in the age range from 80-89 (80’s) years were the other member of the triad for age

groups that scored the highest on laryngeal reflux signs. This age group featured in only

four categories: subglottic oedema, ventricular obliteration, posterior commisure

hypertrophy, and vocal cord oedema. The number of participants in the 50’s age group

was 17,6 in the 70’s and 3 in the 80’s age groups.

Sataloff (2005b) states that changes in true vocal folds related to age include oedema of

the mucosa. Age related physical changes may impact the RFS scores of the said

population. Also, older patients may be taking more medication for concomitant

conditions that may have similar signs to LPR.

93

Conclusion

The results of this research add to the ongoing dispute regarding the diagnosis of

laryngopharyngeal reflux. The results as a whole contribute to the school of thought that

the RSI and the RFS are good predictors of LPR. The significant correlation of the results

although negative, still allows for better predictions on which patients may have LPR,

based on the results of the RSI. The findings essentially afford the team of professionals

the comfort of knowing that the scales do identify LPR signs and symptoms. The total

scores indicate that by examining the RSI a team can with accuracy predict if a patient

will have LPR and also which patients will require further investigation. Clinically the

knowledge derived from this research has impact on the management of patients

presenting at voice clinics and may change the outcome for patients particularly for males

and those with confounding variables.

The results related to associations on intra- and inter-item correlations on the RFS and the

RSI provide further insight into specific symptoms and signs. Intra-item correlation of

the RSI was substantially greater than the intra-item correlation on the RFS. Despite this

both correlations allow us to make inferences of one symptom based on another and

therefore have a better understanding of the various presentations of LPR. It is

unfortunate that between-item correlation of the RSI and the RFS was not possible as this

would have yielded valuable information regarding which symptoms and signs we would

expect to see together. Despite this, the associations of symptoms and signs between the

scales reveal beneficial information on commonly co-occurring factors. Although not

statistically correlated they give clinicians some information on what to expect

particularly at specific severity levels.

Furthermore, issues such as gender and prevalence and severity of LPR were raised. Men

have more severe ratings on the RFS (signs) than females although they typically report

their symptoms as occurring less frequently. This may be due to later disease progression.

Male patients diagnosed with LPR may therefore be at greater risk for laryngeal

carcinoma, oesophageal carcinoma and Barrett’s Disease which are all associated with

prolonged reflux (Sataloff, 2005a). Clinicians should therefore be aware that males may

be at greater risk as they have only decided to seek medical advice and intervention once

the signs are more severe. Therefore although more females may present at a clinic and

94

they may perceive their symptoms as more severe they may be at less risk for carcinoma

than their male counterparts based on the progression of the LPR. One may consider also

that if symptoms resolve more quickly than laryngeal signs (Belafsky et al. 2001a) that

the laryngeal signs may also develop more slowly than the symptoms. If this was the

case, females may have symptoms rated as occurring more frequently or severely despite

their depressed laryngeal findings and this may be an accurate reflection of LPR.

Patients who smoke should also be evaluated with caution as a number of items used to

rate severity of laryngeal signs on the RFS are related to smoking, as is the case with the

RSI symptoms. Smoker’s signs and symptoms may therefore complicate results leading

to an over diagnosis of LPR, especially in heavy smokers who have done so for a number

of years. This group should have full disclosure with their team of specialists regarding

smoking status, time since cessation and classification as a light versus heavy smoker.

This information will have an impact on how the signs are interpreted by the voice team.

One study reported that smoking even only briefly within 5 years prior to assessment can

impact laryngeal signs (Awan & Morrow, 2005).

Professional voice users represent a large proportion of the LPR populations as is

confirmed in this study (non-professional=55 and professional=50). It reinforces the

notion that LPR is prevalent in professional voice users for a number of reasons, but also

raises a few questions. Specifically, questions regarding the nature of a normal

professional voice users’ larynx. Do professional voice users through the nature of their

profession have more frequent symptoms or are they more aware of changes with their

voices as their livelihood depends on it? The fact that a high percentage of healthy

‘normal’ subjects will have at least one sign related to LPR prompts the researcher to

wonder if voice clinics don’t see more professional voice users than non-professional

voice users due to the sensitivity that professional voice users have to subtle changes in

their voice. If this is the case then surely some of these professional voice users may be

similar to the normal healthy non-professional voice users and present with laryngeal

signs of LPR despite being healthy. This should be researched further.

The increased subjective ratings on the RSI by the older population groups indicate the

need for counselling in that population regarding age related voice changes. There are a

number of changes that have been documented in the larynx particularly after the 6th

95

decade (Mortelliti, Malmgren & Gacek, 1990; Kersing & Jennekens, 2004). These

changes may be alarming to this population and misunderstood as to the aetiology. For

this reason counselling that explains the basis for these changes and whether or not

concern is warranted would be beneficial within the clinical setting.

Variables that can impact on the reliability of the RFS and the RSI must be considered.

This is not to say that the RFS and the RSI are not measuring what they purport to do as

this study confirms. However, the scales may not always be able to rule out other

disorders, diseases and factors that may present as LPR. The concern is not whether the

RFS and the RSI diagnose LPR, but rather does the RFS and the RSI over-diagnose LPR.

There is a distinct possibility that this may be true due to the symptoms and signs which

are not exclusive to LPR. Therefore diagnosis of LPR is possible using the scales

evaluated in this research, and there is specificity in the materials utilised to diagnose

LPR. However, the ability to differentially diagnose using these scales is another very

important research project altogether.

The research allows professionals who work with patients who may suffer from LPR to

draw conclusions on laryngeal findings based on reported symptomology. The findings

related to the most prevalent signs and symptoms tie in well with what previous

researchers have found and serve to show that these signs and symptoms should not be

viewed lightly as they indicate a disease for more sinister than its individual signs and

symptoms suggest. It is important to consider that both the subjective RSI and the

objective RFS provide valuable information regarding LPR diagnosis, and further, that

the subjective nature of the RSI cannot be overlooked in LPR diagnosis. Despite the

subjectivity it does in fact provide clinicians with valuable data that is used to diagnose

LPR.

One should also remember that the nature of LPR, with its multitude of symptoms and

signs, does not always fit into a neat package. Although this research provides some

demarcations for where LPR should fit into the picture it does not provide absolutes for

diagnosis based on symptom and laryngeal sign presentation. Instead the current research

should be utilised as a framework that can give some insight into the signs and symptoms

that patients present with and provide a starting point for diagnosis and consequently

correct management. Management is still primarily focused on PPI therapy, diet

96

modification and counselling. One interesting adjunct to symptom management is the use

of chewing gum to alleviate symptoms of reflux. Chewing gum has been found to raise

the pH of the oesophagus and the larynx therefore reducing the symptoms of LPR

(Smoak & Koufman, 2001). The findings also indicated that the buffering effects of the

gum last twice as long as the actual gum chewing itself (Smoak & Koufman, 2001)

Within the literature related to LPR there seems to be a movement away from diagnosis

and more towards management. If a patient responds to the management strategies

employed then diagnosis is confirmed, if not then the hunt for accurate diagnosis

continues. This is the case with trail PPI therapy. This is also the case with a new product

that is in the midst of being patented by Koufman, Ramsey and Battle (patent application

number: 111953,029). The product is a method for treating existing LPR or preventing

further damage to the oesophagus, larynx and hypopharynx from pepsin. A cellulose

powder is inhaled either orally or intranasally. This cellulose powder adheres to the lining

of the areodigestive tracts and becomes a gel on contact with the lining membranes. The

gel prevents pepsin from binding and penetrating the mucosal cells of the areodigestive

tract. They report the cellulose powder used is commonly found in food and cosmetic

products and is non-toxic and indigestible. They state there are no side effects or

overdose complications and that it may be used multiple times a day preventatively or as

symptoms necessitate. They report that preliminary findings suggest that symptoms of

LPR are reduced in over 70% of patients treated. If this method is successful then many

of the issues related to diagnosis will be circumvented and so too will the more severe

complications and consequences of LPR.

Until this time, however, we are bound to try and diagnose LPR effectively using the

methods and materials that are best able to do this accurately and affordably, with the

least patient discomfort. The RFS and the RSI scales are two such methods that this study

has demonstrated show specificity in their diagnosis and can be utilised for such a

purpose.

97

Limitations of the study

Although the relatively large sample size increases the external validity of the study and

the generalisability (leading to a wider application) of results, no provision was made for

socioeconomic status. The clinic that was utilised for the study is a private clinic and

therefore by nature has patients who are of a certain socio-economic status. The results

may thus be difficult to generalise to populations with lower socio-economic status.

In addition, further trends may have been easier to establish if sample sizes were matched

in terms of smoking status, gender and age.

Small sample size for certain age groups may also have skewed the results. The small

sample size for age group 80-89 years (n=3) caused outliers making it appear that many

more participants from that age range had severe symptoms and laryngeal signs, when, in

fact a smaller sample implied larger statistical differences in that age group. Alternatively

these smaller age groups could have been eliminated from the study thereby only

including larger age groups which would have more statistical impact. This however,

would mean little information on the smaller age groups and would only therefore

provide insight into the larger age groups. Further, the older age groups may have scored

higher in their symptom severity based on other extraneous causative variables such as

asthma, diabetes, and cardiac and blood pressure difficulties. Instead this was managed

by including the top three age groups that rated a symptom or laryngeal sign as severe. A

broader group is included in what we consider relevant and the smaller age groups taken

with a measure of caution in interpretation of those results.

Further age related changes in the voice have been documented and changes in the

superior laryngeal nerve showed up to 67% demyelination of these nerve fibres in

patients over 60 years (Mortelliti, Malmgren & Gacek, 1990). These researchers suggest

that these histomorphologic observations may be a direct correlate to the age-related

sensorimotor dysfunction seen in the upper areodigestive tracts of many of the elderly.

Further there have been documented changes in the muscle fibres of the larynx.

Specifically one study found that mitochondrial abnormalities begin to develop from the

6th decade of life in the thyrovocalis muscle and may play a role in the functional deficit

of the larynx in old age (Kersing & Jennekens, 2004). This may have impacted on the

98

results for this age group and therefore it may have been better only to include patients up

to 60 years old in the study.

Patient’s previous medical history, such as previous treatment for GERD, or reflux was

not documented in the study. Therefore of the 105 participants that qualified for the study

it is very possible that a percentage had already undergone some degree of medical

management. It may be that some participants had previously been prescribed a trial

regime of proton pump inhibitors from their general practioner or over the counter reflux

medication. Further, alternative medication may have on effect on the signs and

symptoms that we are measuring with the RFS and the RSI.

Smokers present with a number of the signs related to LPR and used diagnostically on the

RFS. Smokers may therefore have abnormally raised RFS scores on these items

specifically and one should consider excluding patients with even a brief history of

smoking from the sample for a study of this nature. In the study cited in this research

report a smoker was defined as any subject who, at the time of their study, had smoked at

least two cigarettes per day for at least 1 year. Non-smokers were those who, at the time

of their study, did not smoke and who had not smoked for at least 5 years (Awan &

Morrow, 2005). More stringent inclusion criteria and in-depth information into smoking

status of participants should therefore be included and considered in future research.

Implications of the research

The findings of this study will assist with diagnosis and therefore aid in reducing the

number of missed diagnoses of LPR. This has important implications for a patient’s

health. Missed diagnosis translates into increased costs, mental anguish on the patient’s

side, frustration on the clinician’s side and potentially disastrous health implications.

The first step in managing a disease is diagnosis. However, the difficulties that have been

experienced with the diagnosis of LPR have led to a bottleneck when it comes to

management. Poor and imprecise diagnosis has meant that knock-on diseases have

greater impact on the patient. Therefore precise and early detection of LPR is optimal.

Understanding the role that the RFS and the RSI play in diagnosis is crucial. With this

99

greater understanding of how they interact and the knowledge that they have specificity

in diagnosis of LPR, we may in fact be able to streamline the diagnostic process.

Earlier and more accurate diagnosis means that patients may find resolution from their

health queries and quandaries sooner. In a study of the perceived stress, anxiety and

depression in patients with common vocal pathologies it was found that more females

than males had increased levels of anxiety, stress and depression. The population with

vocal pathologies had 25% more stress, 36,9% more anxiety and 31,2% more depression

than healthy controls (Dietrich, Verdolini Abbott, Jackie Gartner-Schmidt,

& Rosen, 2006). In the same study anxiety and stress have been described as causal

triggers, which may in fact exacerbate, and maintain factors that cause voice related

difficulties. This may be true of LPR. The more anxious a patient is about their symptoms

the more severe they perceive their symptoms and perhaps this makes the physical

laryngeal signs worse through depressed immune system functioning and increased

gastrointestinal issues. If LPR is diagnosed timeously the stress on a patient can be

reduced. This in turn will assist at least in part the patients’ recovery as their reduced

anxiety means they can have better recovery outcomes.

Other diseases that are highly associated with LPR, and the presence of which may be

directly caused by LPR can be managed correctly if LPR is accurately diagnosed.

Diseases such as asthma, otitis media, and sinusitis may therefore be better managed and

their effects reduced once LPR is under control. In terms of laryngeal oesophageal

carcinoma and Barrett’s disease, accurate and early diagnosis of LPR will play an

important role in the reduction of such life-threatening offshoots from LPR. Early

detection of such diseases has better prognosis for the patient. Therefore any way in

which it is possible to reduce their effects and increase the patients’ chances of survival

are positive.

By utilising the scales (RFS and RSI) as a mainstay of diagnosis for patients the number

of costly referrals can be reduced. The use of pH - monitoring- which is very expensive

and causes a fair amount of discomfort for the patient can be reserved for the truly

ambiguous case or those with conflicting results. Trial PPI treatment to establish a

diagnosis can be reduced as there is more clarity on those patients who will show more

benefit than others.

100

This study has also thrown light on a population that may be at greater risk for serious

consequences of LPR. Males as discussed only seek intervention later in disease

progression. Therefore with an awareness of this trend clinicians can be more aware of

the pressure for accurate disease diagnosis and correct management of this population.

Further females that present with more severe symptom occurrence on the RSI, yet

reduced laryngeal signs should be viewed in a serious light as these may be the best cases

for early diagnosis if laryngeal signs develop more slowly than symptoms.

In light of the patent pending by Koufman, Ramsey and Battle, the scales as materials

used in LPR diagnosis can be very beneficial. Patients who present with laryngeal signs

and symptoms may be eligible for such treatment which may be a cost effective way of

ruling out those who would benefit and those who would not.

Directions for future research

The discussion and implications for this research has brought to light a few areas that can

be future research questions. The specificity of the RFS and the RSI in diagnosis of LPR

has been established. However, the possibility of over diagnosis still exists. Therefore one

avenue of research should look into the ability of the RFS and the RSI to differentially

diagnose other diseases which have similar laryngeal signs and/or symptoms. If the scales

could be utilised reliably to differentially diagnose other similarly presenting disorders

then they would carry more weight clinically. Factors such as smoking and professional

voice use need to be elaborated on as aspects of what may impact on the accuracy of the

LPR diagnosis. Trends related to professional voice use and smoking status were briefly

examined in this report, but there is scope especially in professional voice use which

seems to comprise such a large sample of voice clinic patients.

It may however, be necessary to redefine the RFS scoring system. Although the RFS does

have clinical significance the awkward scoring means that its items cannot accurately be

correlated against the symptoms on the RSI. A more uniform scoring system for the RFS

should be developed to allow for better intra- and inter-scale validation. Refining the RFS

may aid in differential diagnosis of other similar diseases too.

101

As a follow up research project one could establish what percentage of males compared

to females diagnosed with LPR develop carcinoma. By doing so the effects of the

reported delay in which men seek intervention can be established. Obviously previous

occupational history would have to be eliminated from the study, so that those who

worked with metals, paints, asbestos and in many other construction occupations are not

included in the study. Even if these patients had co-occurring LPR their history would

serve to create confusion on the causative factor of their carcinoma.

An important issue that warrants some investigation is whether or not laryngeal signs

develop more slowly than the symptoms of LPR. This would provide insight into why

certain patients seem to present with many symptoms of LPR yet do not clinically have a

definitive diagnosis of LPR on the RFS. In the defined absence of other similarly

presenting disorders it may be the case that laryngeal signs do develop at a slower pace

than symptoms.

Although the age group from 80-89 years old in this study only comprised a small part of

the sample, they tended to skew data on the RSI specifically towards more severe

occurrences. It may be that these patients are overly cautious regarding their symptoms or

other confounding variables impacted on their scores as well as age related changes of the

larynx. Either way it may be beneficial to include a sample that has specific age ranges,

for example 20-60 years old. Participants within this range had higher means on both the

RFS and the RSI and further, in all likelihood have better health than the older

participants. This would therefore decrease the confounding variables. The chances of

having a larger sample within the age range mentioned is higher as there are fewer older

patients available to participate in research. Further as discussed in the limitations of the

study, it is in patients in their 6th decade of life that start to have laryngeal nerve and

muscular fibre changes and therefore it would be more conducive to eliminate patients

from this age upward in further studies.

There is also at present research into the ability of a pepsin assay to be utilised for LPR

diagnosis. This seems promising and another research option would be to correlate these

findings with the RFS and the RSI results. Assay cultivation is also a fairly non-invasive

102

procedure and may be able to be utilised in unison with the scales instead of pH

monitoring.

The role that hormone changes play on voice particularly in middle aged female patients

is another area that begs examination

“Whereas trends are observed and many clinical practices are accepted widely on

the basis of experience, we need definitive, prospective, evidence-based studies.

Until these studies are produced, critical questions will remain unanswered, and

consistent, optimal patient care will remain elusive”

(Gupta & Sataloff, 2009, p. 147).

103

REFERENCE LIST Abaza, M. M. (2004). Laryngopharyngeal Reflux is different than GERD. Otolaryngology-Head & Neck Surgery, 12(10), 1-3.

Ahmed, T. F., Khandwala, F., Abelson, T. I., Hicks, D.M., Richter, J. E., Milstein, C., & Vaezi, M. F. (2006). Chronic laryngitis associated with gastroesophageal reflux: prospective assessment of differences in practice patterns between gastroenterologists and ENT physicians. Am J Gastroenterol, 101 (3) 470-478.

Ali, M. S. (2008). Laryngopharyngeal Reflux: diagnosis and treatment of a controversial disease. Current Opinion in Allergy and Clinical Immunology, 8(1), 28-33.

Aviv, J. E., Liu, H., Parides, M., Kaplan, S. T., & Close, L. G. (2000). Laryngopharyngeal sensory deficits in patients with laryngopharyngeal reflux and dysphagia. Ann Otol Rhinol Laryngol, 109(11), 1000–1006.

Awan, S. N., & Morrow, D. L. (2005). Videostroboscopic Characteristics of Young Adult Female Smokers vs. Nonsmokers. Journal of Voice, Vol21 (2), 211–223. Ayazi, S., Lipham, J. C., Hagen, J. A., Tang, A. L., Zehetner, J., Leers, J. M., Oezcelik, A., Abate, E., Banki, F., &, DeMeester, S. R. (2010). A New Technique for Measurement of Pharyngeal pH: Normal Values and Discriminating pH Threshold. Journal of Gastrointestinal Surgery, Vol13(8), 1422-1429.

Banks, I. (2001). No man's land: men, illness, and the NHS. BMJ, Vol323, 1058-1060.

Barry, D. W., & Vaezi, M. F. (2010). Laryngopharyngeal reflux: More questions than answers. Cleveland Clinic Journal of Medicine, 77(5), 327-334. Behlau, M. & Oliveira, G. (2009). Vocal hygiene for the voice professional. Current Opinion in Otolaryngology & Head & Neck Surgery, 17(3), 149-154. Belafsky, P. C., Postma, G. N., Amin, M. R., Koufman, J. A. (2002). Symptoms and findings of laryngopharyngeal reflux. Ear Nose Throat Journal, Vol81(suppl2), 10-13.

104

Belafsky, P. C., Postma, G. N., & Koufman, J. A., (2001a). Laryngopharyngeal reflux symptoms improve before changes in physical findings. Laryngoscope, 111, 979–981. Belafsky PC, Postma GN, Koufman JA. (2001b). The validity and reliability of the reflux findings score (RFS). Laryngoscope, 111, 1313–1317. Belafsky, P. C., Postma, G. N., & Koufman, J. A. (2002). Validity and Reliability of the RSI. Journal of Voice, Vol16, (2), 274-277. Belafsky, P. C., Rees, C. J., Rodriguez, K., Pryor, J. S., & Katz, P. O. (2008). Esophagopharyngeal reflux. Otolaryngology-Head and Neck Surgery, 13(1), 57-61. Bove, M. J. & Rosen, C. (2006). Diagnosis and Management of laryngopharyngeal reflux disease. Current opinion in Otolaryngology Head and Neck Surgery, Vol3. 116-123.

Branski, R. C., Bhattacharyya, N., & Shapiro, J. (2002). The reliability of the assessment of endoscopic laryngeal findings associated with laryngopharyngeal reflux disease. Laryngoscope, 112, 1019-1024.

Carrau, R. L., Khidr, A., Gold, K. F., Crawley, J. A., Hillson, E. M, Koufman, J. A., & Pashos, C. H. (2005). Validation of a Quality-of Life Instrument for Laryngopharyngeal Reflux. Archives Otolaryngology Head Neck Surgery, Vol13, 315-320. Chen, S. H., Chiang, S-C., Chung, Y-M., Hsiao, L. C., & Hsiao, T. Y. (2010). Risk Factors and Effects of Voice Problems for Teachers. Journal of Voice, 24(2), 183-192. Cohen, J.T.; Bach, K.K.; Postma, G.N., & Koufman, J.A. (2002). Clinical manifestations of laryngopharyngeal reflux. Ear Nose & Throat Journal, 81(9), 19-23 Cohen, J. T., Gil, Z., & Fliss, D., M. (2005). The Reflux symptom Index a clinical tool for the diagnosis of Laryngopharyngeal reflux. Harefuah, Vol144(12), 826-829. Cohen, S. M., & Garrett, C., G. (2008). Hoarseness: Is It Really Laryngopharyngeal Reflux? Laryngoscope 118, 363-366. Colton, R. H. & Casper, J. K. (1996). Understanding Voice Problems a Physiological Perspective for Diagnosis and Treatment (2nd ed.). Baltimore: Williams and Wilkins.

105

Cukier-Blaj, S., Bewley, A., Aviv, J. E., & Murry, T. (2008). Paradoxical Vocal fold Motion. A Sensory-Motor Laryngeal Disorder. The Laryngoscope, Vol118, 367-371. Dallal, G. E. (2003). Why P=0, 05? Retrived January 29, 2011, from http://www.tufts.edu/~gdallal/p05.htm Dietrich, M., Verdolini Abbott, K., Gartner-Schmidt, J., & Rosen, C. A. (2006). The Frequency of Perceived Stress, Anxiety, and Depression in Patients with Common Pathologies Affecting Voice. Journal of Voice, Vol22(4), 472–488. Doustmohammadian N, Naderpour M, Khoshbaten M, Doustmohammadian A. (2010). Is there any association between esophagogastric endoscopic findings and laryngeal cancer? Am J Otolaryngol, Nov, doi:10.1016/j.amjoto.2010.09.013 Eller, R., Ginsburg, M., Lurie, D., Heman-Ackah, Y., Lyons, K., & Sataloff, R. (2009). Flexible Laryngoscopy: A Comparison of Fiber Optic and Distal Chip Technologies—Part 2: Laryngopharyngeal Reflux. Journal of Voice, Vol23 (3), 389-395. Ford, C. N. (1999). Advances and refinements in phonosurgery. Laryngoscope, Vol109, 1891-1900. Ford, C. N. (2005). Evaluation and Management of Laryngopharyngeal Reflux. JAMA, Vol294, 1534-1540. Fouad, Y. T. & Rifaat, A. M. (2010). Treatment of Clinically Diagnosed Laryngopharyngeal Reflux Disease. Archives of Otolaryngology -- Head & Neck Surgery, 136(11), 1089-1092. Fritzell, B., & Hertegard, S. (1986). A retrospective study of treatment for vocal fold edema: a preliminary report. In J. A. Kirchner (Ed), Vocal Fold Histopathology: A Symposium. (pp 57-61). San Diego, CA: College-Hill. Galli, J., Cammarota, G., Volante, M., De Corso, E., Almadori, G., & Paludetti, G. (2006). Laryngeal carcinoma and laryngo-pharyngeal reflux disease. Acta Otorhinolaryngol Ital, 26(5), 260–263.

Gugatschka, M., Schoekler, B., Kiesler, K., & Friedrich, G. (2008). Do clinical Symptoms and laryngoscopic findings of laryngo-pharyngeal reflux correlate. Laryngo- rhino –otologie. Vol 87 (12). 867-869.

106

Gupta, R & Sataloff, R. T. (2009). Laryngopharyngeal reflux: current concepts and questions. Current Opinion in Otolaryngology & Head & Neck Surgery, Vol17(3). 143-148. Halum, S. L., Postma, G. N., Johnston, C., Belafsky, P. C. & Koufman, J. A. (2005). Patients with Isolated Laryngopharyngeal Reflux are not Obese. The Laryngoscope, Vol115, 1042–1045.

Hammer, H. F. (2009). Reflux – associated laryngitis and Laryngopharyngeal reflux: a gastroenterologist’s point of view. Dig Dis, 27(1), 14-17. Hanson, D. G. & Jiang, J., J. (2000). Diagnosis and Management of chronic laryngitis associated with reflux. American Journal of Medicine, Vol108 (Suppl 4a), 112-119. Heman-Ackah, Y. D., Dean, C. M., & Sataloff, R. T. (2002). Strobovideolaryngoscopic Findings in Singing Teachers. Journal of Voice Vol16 (1), 81–86. Hicks, D. M., Ours, T. M., Abelson, T. I., Vaezi, M. F., & Richter, J. E. (2002).The prevalence of hypopharynx findings associated with gastroesophageal reflux in normal volunteers. J Voice 16(4), 564–579 Iwao Honjo, I., & Isshiki, N. (1980). Laryngoscopic and Voice Characteristics of aged Persons. Arch Otolaryngol. Vol106 (3), 149-150. Howell, D. C. (1999). Fundamental statistics for the behavioural Sciences. (4th ed.). California: Brooks/Cole Publishing Company. Jacobson, B. H., Johnson, A., Grywalski, C., Jacobson, G., Benninger, M. S., & Newman, C. W. (1997). The Voice Handicap Index: development and validation. American Journal of Speech-Language Pathology, Vol6, 66-70. Jin, B. J., Lee, Y. S., Jeong, S. W., Jeong, J. H., Lee, S. H., & Tae, K. (2008). Change of Acoustic Parameters Before and After Treatment in Laryngopharyngeal Reflux Patients. Laryngoscope, 118(5), 938-941 Joniau, S., Bradshaw, A., Esterman, A., 7 Carney, A., S. (2007). Reflux and Laryngitis: A Systematic review. Otolaryngology-Head and Neck Surgery, Vol136, 686-692.

107

Johnston, N., Knight, J., Dettmar, P. W., Lively, M.O., & Koufman, J. (2004). Pepsin and carbonic anhydrase isoenzyme III as diagnostic markers for laryngopharyngeal reflux disease. Laryngoscope, 114, 2129–2134.

Karkos, P., & Wilson, J. A. (2006). Empiric Treatment of Laryngopharyngeal Reflux with Proton Pump Inhibitors: A Systematic Review. Laryngoscope, 116, 144–148.

Kay Elemetrics 9100B rhino- laryngeal stroboscope.

Kersing, W., & Jennekens, F. G. I. (2004). Age-related changes in human thyroarytenoid muscles: a histological and histochemical study. European Archives of Oto-Rhino-Laryngology, Vol261 (7), 386-392.

Khan, A. M., Hashmi, S. R., Elahi, F., Tariq, M., & Ingrams, D. R. (2006). Laryngopharyngeal reflux: A Literature Review. Surgeon, 4(4), 221-225.

Kotby, M. N., Hassan, O., El-Makhzangy, A. M. N., Farahat, M., Shadi, M., & Milad, P. (2010). Gastroesophageal reflux/laryngopharyngeal reflux disease: a critical analysis of the literature. Eur Arch Otorhinolaryngol, 267, 171–179.

Koufman, J. A. (1991). The otolaryngologic manifestations of gastroesophageal reflux disease (GERD): a clinical investigation of 225 patients using ambulatory 24-hour pH monitoring and an experimental investigation of the role of acid and pepsin in the development of laryngeal injury. Laryngoscope 101, (53), 1–78.

Koufman, J. A., Amin, M. R., & Panetti, M. (2000). Prevalence of reflux in 113 consecutive patients with laryngeal and voice disorders. Otolaryngol Head Neck Surg 123, 385-388.

Koufman, J. A., Aviv, J. E., Casiano, R. R., & Shaw, G. Y. (2002). Laryngopharyngeal Reflux: Position statement of the Committee on speech, Voice, and Swallowing disorders of the American academy of Otolaryngology- Head and Neck Surgery. Otolaryngology-Head and Neck Surgery, Vol27(1), 32-35.

108

Lee, B.E, Kim, G. H., Ryu, D, Y., Kim, D, U., Cheong, J, H., Lee, D. G., & Song, G. A. (2010). Combined Dual Channel Impedance/pH-metry in Patients With Suspected Laryngopharyngeal Reflux. J Neurogastroenterol Motil, 16(2), 157–165. Lindstrom, D. R., Wallace, J., Loehrl, T. A., Merati, A. L., & Toohill, R. J. (2002). Nissen Fundoplication Surgery for Extraesophageal Manifestations of Gastroesophageal Reflux (EER). The Laryngoscope, 112, 1762–1765.

Maertens, K., de Jong, F. I. (2007). The voice Handicap Index as a tool for the assessment of biopsychosocial impact of voice problems. B-ENT,3(2), 61-66. Martin, S. (1987). Working with Dysphonics A Practical Guide to Therapy for Dysphonia. :Winslow Press. Maron, L. & Jordaan, H. (2010). Gastro-oesophageal reflux disorder (GORD) vs. laryngo-pharyngeal reflux (LPR). The specialist forum, 10(10), 47-52. Mainie, I., Tutuian, R., Shay, S., Vela, M., Zhang, X., Sifrim, D., & Castell, D. O. (2006). Acid and non‐acid reflux in patients with persistent symptoms despite acid suppressive therapy: a multicentre study using combined ambulatory impedance‐pH monitoring. Gut, 55(10), 1398-1402. Mesallam, T. A., Stemple, J. C., Sobeih, T. M.,& Elluru, R. G. (2007). Reflux Symptom Index Versus Reflux Finding Score. Annals of Otology, Rhinoiogy & Laryngology 1 (6), 436-440. Miller MK, Verdolini K. (1995). Frequency and risk factors for voice problems in teachers of singing and control subjects. J Voice. 9, 348–362. Milstein, C., Charbel, S., Hicks, D. M., Abelson, T. I., Reicheter, J. M., & Vaezi, M. F. (2005). Prevalence of laryngeal irritation signs associated with reflux in asymptomatic volunteers: impact of endoscopic technique (rigid vs. flexible scope). Laryngoscope, 115(12), 2256-2261. Morrison, M.; Nichol, H., & Rammage, L. (2001). Management of the Voice and its Disorders (2nd ed.). Canada: Singular publishing

109

Mortelliti, A. J., Malmgren, L. T., & Gacek, R. R. (1990). Ultrastructural Changes With Age in the Human Superior Laryngeal Nerve. Arch Otolaryngol Head Neck Surg, 116(9), 1062-1069.

Mosca, F., Rossillo, V., & Leone, C. A. (2006). Manifestations of gastro-pharyngo-laryngeal reflux disease. Acta Otorhinolaryngol Ital, 26(5), 247–251. Muderris, T., Gokcan, M. K., & Yorulmaz, I. (2009). The Clinical Value of Pharyngeal pH Monitoring Using a Double-Probe, Triple-Sensor Catheter in Patients With Laryngopharyngeal Reflux. Arch Otolaryngol Head Neck Surg, 135(2). 163-167. Murry, T., Tabaee, A., Aviv. J. E. (2004). Respiratory Retraining of Refractory cough and Laryngopharyngeal Reflux in Patients with Paradoxical Vocal Fold Movement Disorder. The Laryngoscope, 14, 1341-1345.

Musser, J., Kelchner, L., Neils-Strunjas, J., & Montrose, M. (2009). A Comparison of Rating Scales Used in the Diagnosis of Extraesophageal Reflux. Journal of Voice. doi:10.1016/j.jvoice.2009.11.009

Noordzij, J. P., Khidr, A., Desper, E., Meek, R.,B., Reibel, J., F., Levine, P., A. (2002). Correlation of pH probe-measured Laryngopharyngeal relux with symptoms and signs of reflux laryngitis. Laryngoscope, Vol112(2), 2192-2195. Oelschlager, B. K., Eubanks, T. R., Oleynikov, D., Pope, C., Pellegrini, C. A. (2002). Symptomatic and physiologic outcomes after operative treatment for extraoesophageal reflux. Surgery Endosc, Vol16, 1032-1036. Orlikoff, R. F. & Baken, R., .J. (1993). Clinical Speech and Voice Assessment. San Diego: Singular Publishing Papakonstantinou, L, Leslie, P., Gray, J., Chadwick, T., Hudson, M., & Wilson, J. A. (2009). Laryngopharyngeal reflux: a prospective analysis of a 34 item symptom questionnaire. Clinical Otolaryngology, 34(5), 455-459.

Phyland, D. J., Oates, J.,& Greenwood, K. M. (1999). Self-reported voice problems among three groups of professional singers. J Voice 13, 602–611.

110

Postma, G. N., & Halum, S., L. (2006,). Laryngeal and pharyngeal complications of gastroesophageal reflux disease. GI Motility online. May, doi:10.1038/gimo46

Pribuisiene, R., Uloza, V., Jonaitis. (2002). Typical and atypical symptoms of laryngopharyngeal reflux disease. Medicina, vol38(7), 699-705. Printza, A., Speletas, M., Triaridis, S., & Wilson, J. (2007). Is pepsin detected in the saliva of patients who experience pharyngeal reflux? Hippokratia, 11(3), 145–149. Punch, K. F. (2005). Introduction to Social Research – Quantitative and Qualitative Approaches. London: Sage Publications. Qadeer, M. A., Swoger, J., Milstein, C., Hicks, D. M., Ponsky, J., Richter, J.E., Abelson, T.I., & Vaezi, M.F. (2005). Correlation between symptoms and laryngeal signs in laryngopharngeal reflux. Laryngoscope, 115(11), 1947-1952. Randhawa, P. S., Mansuri, S., & Rubin, J. S. (2010). Is dysphonia due to allergic laryngitis being misdiagnosed as Laryngopharyngeal reflux? Logoped Phoniatr Vocol, 35(1),1-5. Reimer, C. & Bytzer, P. (2008). Management of Laryngopharyngeal Reflux with Proton Pump Inhibitors. Ther Clin Risk Manag, vol4(1), 225-233. Remacle, M., & Lawson, G. (2006). Diagnosis and management of Laryngopharyngeal reflux disease. Current Opinion in Otolaryngology & Head & Neck Surgery, Vol12(3), 143-149. Rosen, C. A., Lee, A. S., Osborne, J., Zullo, T., & Murry, T. (2004). Development of the Voice Handicap Index-10. Laryngoscope, 114(9), 1549-1556. Rouev, P., Chakarski, I., Doskov, D. Dimov, G., & Staykova, E. (2004). laryngopharyngeal Symptoms and Gastroesophageal Reflux Disease. Journal of Voice, Vol19(3), 476–480. Roy, N., Merrill, R. M., Gray, S. D., & Smith, E. M. (2005). Voice disorders in the general population: Prevalence, risk factors, and occupational impact. The Laryngoscope, Vol115(11), 1988-1995.

111

SAS 9, 2 Program Sataloff, R. T. (2005a). Clinical Assessment of Voice. San Diego: Plural Publishing. Sataloff, R. T. (2005b). Treatment of Voice Disorders. San Diego: Plural Publishing. Sataloff, R. T, Castell, D. O., Katz, P. O., & Sataloff, D. M. (2006). Reflux Laryngitis and Related Disorders (3rd ed.). San Diego, California: Plural Publishing. Sataloff, R. T., Hawkshaw, M. J., & Gupta, R. (2010). Laryngopharyngeal reflux and voice disorders: an overview on disease mechanisms, treatments, and research advances. Discov Med, 10(52), 213-24. Schiavetti, N., & Metz, D. E. (2002). Evaluating Research in Communication Disorders. (4th ed.). Boston: Allyn and Bacon. Schindler, A., Mozzanica, F., Ginocchio, D., Peri, A., Bottero, A., & Ottaviani, F. (2010). Reliability and Clinical Validity of the Italian Reflux Symptom Index. Journal of Voice, 24(3), 354-358. Schreiber, S., Garten, D., & Sudhoff, H. (2009). Pathophysiological mechanisms of extraesophageal reflux in otolaryngeal disorders. Eur Arch Otorhinolaryngol, 266, 17-24. Siupsinskiene, N., Adamonis, K., & Toohill, R. J. (2007). Quality of life in laryngopharyngeal reflux patients. Laryngoscope, 117(3), 480-484. Smoak, B. R., & Koufman, J. A. (2001). Effects of gum chewing on Pharyngeal and Esophageal pH. Annals of Otlology, Rhinology and Laryngology, 110(12), 1117-1119.

So, J. B., Zeitels, S. M., & Rattner, D. W. (1998). Outcome of atypical symptoms attributed to gastroesophageal reflux treated by laparoscopic fundoplication. Surgery, 124, 28–32.

Srinivasan, R., Vela, M. F., Katz, P.O., Tutuian, R. Castell, J. A., & Castell, D. O. (2001). Esophageal function testing using multichannel intraluminal impedence. American Journal of Physiology of Gastrointerology and liver physiology, Vol28, 457-462.

112

Tanner, D. C. (2007). Medical-Legal and Forensic Aspects of Communication Disorders, Voice Prints, and Speaker Profiling. Arazona: Lawyers & Judges Publishing Company. Thibeault, S. L., & Zelazny, S. K. (2010). Assessment of hoarseness and dysphonia. Retrieved January 17, 2011 from http://bestpractice.bmj.com/best-practice/monograph/845/diagnosis.html Titze, I. R. (1994). Principles of Voice Production. New Jersey: Prentice-Hall.

Ulualp, S. O., Toohill, R. J., & Shaker, R. (2001) Outcomes of acid suppressive therapy in patients with posterior laryngitis. Otolaryngol Head Neck Surg, 124, 16–22.

Ulis, J.M., & Yanagisawa, E. (2009). What's new in differential diagnosis and treatment of hoarseness? Current Opinion in Otolaryngology & Head & Neck Surgery, 17(3), 209-215.

Vaezi, M. F., Hicks, D. M., & Abelson, T. L. (2003) Laryngeal Signs and Symptoms and gastroesophageal reflux disease (GERD): a critical assessment of cause and effect association. Clinical Gastroenterology Hepato, Vol1, 333-344.

Vázquez de la Iglesia, F., Fernández González, S., & Gómez Mde, L. (2007). Laryngopharyngeal reflux: correlation between symptoms and signs by means of clinical assessment questionnaires and fibroendoscopy. Is this sufficient for diagnosis? Acta Otorrinolaringol Esp, 58(9), 421-425.

Verduyckt, I., Remacle, M., Jamart, J., Benderitter, C.,& Morsomme, D. (2009). Voice Related Complaints in the Paediatric Population. Journal of Voice. Article in press. doi:10.1016/j.jvoice.2009.11.008

Wang, L., Liu, X., Liu, Y. L., Zeng, F. F., Wu, T., Yang, C.L., Shen, H. Y., & Li, X. P. (2010). Correlation of pepesin-measured laryngopharyngeal reflux disease with symptoms and signs. Otolaryngology - Head and Neck Surgery, 143(6), 765-771. Welham, N. V., Dailey, S. H., Ford, C. N., & Bless, D. M. (2007). Voice Handicap Evaluation of Patients with Pathologic Sulcus Vocalis. Anitals of Otoligy, Rhinology & Laryngology, 116(6), 411-417.

113

Welman, J. C., & Kruger, S. J. (1999). Research Methodology for the Business and Administrative Sciences. Cape Town: Oxford University Press. Westcott, C. J., Hopkins, M. B., Bach, K., Postma, G. N., Belafsky, P. C., Koufman, J. A., (2004).Fundoplication for laryngopharyngeal reflux disease .Journal of the American College of Surgeons, Vol199(1), 23-30. Williams, R. B., Szczensniak, M. M., & Maclean, J. C., (2004). Predictors of outcome in an open label, therapeutic trial of high-dose omerprazole in laryngitis. American Journal Gastroenterology, Vol99, 777-785. .

114

Appendix A

CONSENT FOR THE USE OF CLINICAL INFORMATION

This document must be explained to the patient by a member of the clinical staff

You are presently a patient or are scheduled to be seen for problems you are currently experiencing at the Wits University Donald Gordon Voice and Swallowing Centre. The centre not only renders treatment but is also actively involved in conducting research aimed at improving the quality of the care we deliver. From time to time such research involves the use of patient records from which information is extracted. The use of such information is subject to

• Approval from the Committee for Research on Human Subjects (University of the Witwatersrand).

• Anonymity: in other words the identity of the patient from whose file information is extracted is never revealed to anyone but the researcher unless specific consent is obtained from the patient to do so.

We would like to obtain your consent to use information from you / the patient’s file for the purpose of our research, subject to the above mentioned conditions. If you choose not to give your consent, then doing so will not compromise your or the patient’s current or future treatment in any way. If at any time in the future, you choose to withdraw this consent, you are free to do so and this again will not prejudice your or the patient’s current or future treatment in any way. Should you wish to contact us at any stage regarding this consent, please contact the voice and swallowing centre on (011) 482 5524 ______________________________ ____________________________ Print patient’s full name and surname VSC File Number (for office use)

DELETE WHICHEVER IS NOT APPLICABLE

YES I the undersigned, hereby give consent for my / the patient’s records to be used as per the above mentioned conditions for the purpose of research. NO I do not give consent for the use of my / the patient’s records for the purpose of research __________________________________ ___________________________ Full name of person giving/declining consent Write ‘SELF’ or give relationship to patient Signed at ___________________ on _______________________________ ____________________________ ____________________________ Patient Person Giving Consent

115

Witness #1 _________________________ Signature: _____________________

116

Appendix B

RSI

Within the last month how did the following problems affect you? 0= No problem 5= Severe problem

1. Hoarseness or a problem with your voice. 0 1 2 3 4 5 2. Clearing your throat. 0 1 2 3 4 5 3. Excess mucous or postnasal drip. 0 1 2 3 4 5 4. Difficulty swallowing food, liquids or pills. 0 1 2 3 4 5 5. Coughing after you ate or after lying down. 0 1 2 3 4 5 6. Breathing difficulties or choking episodes. 0 1 2 3 4 5 7. Troublesome or annoying cough. 0 1 2 3 4 5 8. Sensation of something sticking or a lump in your throat 0 1 2 3 4 5 9. Heartburn, chest pain, indigestion or stomach acid coming up. 0 1 2 3 4 5

117

Appendix C

Reflux Finding Score

Finding Possible Score Actual score

Subglottic Oedema 0 = absent 2 = Present

Ventricular Obliteration 2= Partial 4 = Complete

Erythemia/ hyperemia 2 = Arytenoids only 4 = Diffuse

Vocal Cord Oedema 1 = Mild 2=Moderate 3= Severe 4 =Polypoid

Diffuse Laryngeal Oedema 1= Mild 2= Moderate 3= Severe 4=Obstructing

Posterior commissure

hypertrophy 1=Mild 2=Moderate 3=Severe 4= Obstructing

Granuloma/ granulation 0 = absent 2=Present

Thick endolarngeal mucus

/Other 0 = absent 2=Present

Total Reflux Finding Score

118

Appendix D

Letter from the site providing permission to conduct research

119

Appendix E

Ethical Clearance Certificate